A simple technique for characterizing flexible materials using pressurised tubes Authors: Manuel Rubio, Joaquín Anatol, César Barrios-Collado, Miguel Ángel Herrada Published: 2026 ABSTRACT AbstractThis work numerically and experimentally investigates the deformation of flexible tubes under positive transmural pressure. The numerical analysis employs asymmetric static simulations using a neo-Hookean model defined by the shear modulus, and is validated through experiments with two different materials. A dimensionless study was conducted to evaluate the variation in tube volume as a function of transmural load (the ratio of transmural pressure to shear modulus), considering different lengths and thicknesses. For sufficiently slender tubes, reducing the thickness by one order of magnitude requires a transmural load smaller by the same order to achieve the same deformation. Additionally, tube length does not affect deformation for relatively long tubes. Therefore, for tubes that are both sufficiently long and slender, the volumetric deformation as a function of the ratio between transmural load and dimensionless thickness collapses onto a single curve, simplifying deformation prediction. This curve enables the shear modulus to be estimated from a single experiment using a manometer, facilitating material characterization.Keywords: Flexible tubes, Transmural pressure, Shear modulus, Neo-Hookean modelSubject terms: Engineering, Materials science, Physics Introduction Numerous industrial and biomedical applications employ flexible tubes that interact with the flow of internal fluid. Examples of industrial applications include membrane infiltration processes1, elastomeric infusion pumps2, or Liebau pumping3,4. In the biomedical context, the materials used for artificial blood vessels5, heart valves6 or stents7 must replicate the elastic properties of biological tissues. Moreover, this phenomenon also appears in human physiology, such as in pulmonary airways, large blood vessels, the urethra, and other anatomical structures. This fluid–structure interaction is responsible for pulmonary maximal forced expiration and for the generation of Korotkoff sounds. These phenomena have led to a line of research involving the experimental analysis of Starling resistors8, in which the relationship between transmural pressure (the difference between internal and external pressures in the tube), flow and tube deformation was examined. Under positive transmural pressure, the tube simply reduces its cross-sectional area in an axisymmetric manner. However, under negative transmural pressure, the tube flattens until the walls come into contact, collapsing in a non-axisymmetric fashion. Under certain conditions, the tube exhibits self-excited oscillations9,10. This complex fluid–structure interaction is three-dimensional and has been studied by modeling the solid as a thin shell11, with the aim of reducing the computational cost of the problem. Another possibility is to study it in a simplified manner using one-dimensional models, in which the equations governing the motion of the tube are replaced by a “tube law” that relates transmural pressure to the cross-sectional area of the tube12,13. It has also been investigated by modeling the solid as a thick-walled structure14,15. Accurate structural modeling is of fundamental importance in all these studies. Accordingly, a substantial body of research has been devoted to the mechanical behavior and potential instabilities of flexible tubes subjected to pressure, even in the absence of internal flow. In particular, the static deformation of cylindrical tubes under positive transmural pressure has been extensively investigated within the framework of nonlinear elasticity. Classical analytical contributions, beginning with the work of Rivlin16 and further developed in studies such as Alexander17, Chadwick18,19, Haughton20,21, and later synthesized by Ogden22, established the fundamental solutions for the inflation of thin- and thick-walled hyperelastic tubes. In this regime, the deformation remains axisymmetric and is characterized by a monotonic relationship between the internal pressure and the radial expansion, governed by the constitutive behavior of the material. These solutions show that, for positive transmural pressure, the tube undergoes smooth and stable expansion while preserving its cylindrical symmetry. Complementary analyses based on membrane and shell theories23–25 further confirm that, in the absence of flow-induced effects and instability mechanisms, the structural response is fully determined by geometric and material parameters, as well as by boundary conditions, without the emergence of non-axisymmetric deformation modes. Post-bifurcation analyses for hyperelastic tubes investigate the role of bending stiffness in the initiation and evolution of localized aneurysm-type deformations26,27. Extensions to residually stressed and biomechanically motivated materials have also been developed on inflation, extension, and torsion of residually stressed tubes28; and the broader constitutive and stability framework for soft tissues29. The works on externally pressurized thin shells focus on structural resistance, which is generally governed by buckling instability rather than material yield. These structures exhibit a highly nonlinear mechanical response strongly influenced by geometric imperfections, slenderness of the shell, boundary conditions, and residual stresses, so real shells can collapse at pressures significantly below the theoretical critical load. This was originally established in the classical elastic stability formulations developed by Timoshenko and Gere24 and the imperfection-sensitivity theory introduced by Koiter30. Subsequent experimental and analytical investigations confirmed the decisive role of geometric imperfections and nonlinear prebuckling deformations in reducing the collapse capacity31–33. Other studies have focused on nonlinear finite element approaches, stochastic imperfection modeling, and refined knock-down factors to improve the prediction of realistic collapse pressures under hydrostatic loading34–38. In all of these approaches, the mechanical constitutive law of the flexible material plays a crucial role. Numerous material models or constitutive models are used to define the constitutive law of the tube. Among hyperelastic models, the neo-Hookean model is noteworthy; it employs the shear modulus (G) as its parameter, which relates the resulting deformation to the applied shear stress and can be used to predict the nonlinear stress–strain response of materials undergoing large deformations39. The constitutive laws of viscoelastic materials are generally based on virtual chains of indefinite length whose links consist of elementary springs and dampers. The most widely used models include the Maxwell, Kelvin–Voigt, and Standard Linear Solid models. To characterize materials such as tissues or biogels, it is common to use dynamic moduli, known as the storage modulus (G') and the loss modulus (G'')40. This characterisation is performed through oscillatory rheological tests, using a rheometer to analyse the dynamic deformation of the material when a sinusoidal torsional stress is applied over a wide range of frequencies41. These moduli can then be used to determine the parameters of the viscoelastic models mentioned above. However, specialised equipment is required to obtain them. Therefore, having a simple method for characterising such materials, one that provides a parameter directly used in the models–such as the shear modulus–would represent a step forward. This work addresses the characterization of the shear modulus in thick flexible tubes subjected to positive transmural pressure, both experimentally and numerically, under static conditions. To this end, the study is divided into three sections. Section 2, devoted to methods, describes the configuration under investigation, the governing equations of the phenomenon, and the numerical and experimental methodologies employed. Section 3, on discussion and results, validates the numerical model and subsequently conducts a dimensionless parametric study. In addition, this section provides an expression that predicts the volume of a hyperelastic flexible tube as a function of transmural pressure in a dimensionless form. Finally, Section 4 presents the conclusions of the study. Methods Definition of the geometrical domain Figure 1 shows the geometrical domain studied in this work, in which a flexible tube with circular cross-section undergoes axisymmetric deformation. The tube is attached at both ends to two ideally rigid tubes. As the same liquid fills the domain, both inside and outside the tubes, albeit at different pressures ( and respectively),the effect of body forces (gravity) is minimised, and thus can be neglected. So, the transmural pressure is defined as Δ p = -. When the transmural pressure is null, the diameter, length, and thickness of the flexible tube are indicated as d, l, and e, respectively (Fig. 1-top). In this case, the geometrical characteristics are uniform and the internal volume is = . In contrast, when positive transmural pressure is applied, the volume, V, increases (Fig. 1-bottom) and the diameter and thickness vary along the axial direction of the tube. Geometrical domain; (top) null transmural pressure; (bottom) positive transmural pressure. Under static transmural pressure, the deformation of the flexible tube is determined according to its material mechanical properties, so constitutive laws such as neo-Hookean models can be applied. As section 2.2 explains, the shear modulus (G) will be the only mechanical property to be considered. The problem is then reduced to the study of the deformed volume, V, as a function of reference dimensions (d, l and e), a material property (G) and an operating condition (Δ p). Mathematical model The equations governing mechanical stresses in a flexible tube under the axisymmetric hypothesis are the following. Applying a neo-Hookean solid model, stresses are determined as a function of pressure p and displacements at radial and axial directions: The tube is assumed to be incompressible, so the condition defined in42 is imposed, The boundary conditions for the radial stresses at the inner () and outer () walls of the flexible tube are determined by the transmural pressure, Finally, since at the ends of the flexible tube, the displacements must be null, since it is attached to rigid tubes, so and . Numerical method The numerical model is based on Herrada & Montanero 43 for interfacial flows and was further extended by Snoeijer et al. 42 to account for hyperelastic solids. The spatial domain for the solid is mapped onto a rectangular domain using Cartesian coordinates r and z (Fig. 1). The governing equations are discretized using a Chebyshev collocation method in the r-direction, employing collocation points. Spatial derivatives in this direction are computed using spectral differentiation matrices. In the z-direction, derivatives are approximated using fourth-order finite difference schemes on a uniform grid with points. The resulting nonlinear system of algebraic equations is assembled and solved using the freely available JAM (Jacobian Analytical Method44) framework developed by Herrada and implemented in MATLAB. Within this framework, the nonlinear system is solved using a Newton–Raphson iterative procedure, in which the Jacobian matrix is automatically constructed following the JAM methodology, avoiding explicit analytical derivation while preserving the quadratic convergence of the method. Further details on the JAM method can be found in45. In this work, the linear system at each Newton iteration is solved using MATLAB built-in routines, taking advantage of the sparse structure of the Jacobian matrix. Boundary conditions are enforced directly at the collocation points by replacing the corresponding equations in the global system. Convergence is achieved when the norm of the solution update or residuals falls below a prescribed tolerance. The procedure followed in each simulation is always the same. First, the mesh is generated under zero transmural pressure conditions, for which the conduit has a reference volume . Subsequently, the transmural pressure, Δ p, is gradually increased, and the resulting changes in shape and volume, V, are computed. The results of this procedure are illustrated in Fig. 2 for two different meshes. As can be observed, the results are converged for a mesh with and . To ensure an adequate axial resolution for different lengths, the number of nodes is adjusted according to the relation . Volume–pressure relationship of the flexible tube for two different meshes (d = 10 mm, l = 100 mm, e = 1 mm, G = 1 MPa). The inset shows the mesh with and .. The numerical method employed is computationally efficient. This is mainly due to the assumption of axisymmetry, which reduces the problem to two dimensions. However, this assumption limits its applicability to situations where axisymmetry holds. One example that cannot be simulated is the case of negative transmural pressure, as the resulting tube shape is no longer axisymmetric12,30,46. For this reason, only positive transmural pressures are considered in the present study. Experimental set-up The experimental set-up (Fig. 3) consisted of a cubic methacrylate reservoir fitted with two aligned bulkhead connections on opposing faces, where both ends of the flexible tube are attached. A worm-screw-driven syringe pump injected known volumes of liquid into the flexible tube, and the inner pressure was measured by means of a manometer (Eirelec 5000PUH, Ireland). Table 1 shows the characteristics of the two tested flexible tubes, in-house made latex47 and vulcanised rubber. Water was used as both the inner and the outer fluid. The uncertainties for the pressure and volume measurements were, respectively, 500 Pa and 1 ml. Image (top) and sketch (bottom) of the experimental set-up. Table 1Characteristics of the tested flexible tubes.Flexible tubeLatexRubberd20 mm19 mme1 mm0.9 mml136 mm136 mm42.7 ml38.6 ml Characteristics of the tested flexible tubes. The experimental procedure consisted of the following steps. First, the flexible tube was attached to the set-up, and the reservoir was filled until the free surface rose above the tube to neglect body forces. Then, the flexible tube was filled by means of the syringe pump until the reference state was achieved ( and Δ p = 0). Afterwards, air was purged. Finally, pressures were recorded for a set of volumes ranging from until the apparition of non-axisymmetric deformation (noted as large, localised bulges). In our experiments, such deformations are observed to arise when , in agreement with the literature26. In contrast, throughout the present study we have at all times, and therefore no unstable regime is expected to occur. For each tube tested, three series of pressurizing and depressurizing cycles were performed, observing that hysteresis was negligible. Results and discussion Validation and experimental characterization The validation will consist of three stages. First, a uniaxial tensile test will be performed under zero transmural pressure to assess whether the neo-Hookean model adequately describes the material behavior. Subsequently, the tubes will be subjected to transmural pressure, and an experiment–simulation comparison will be carried out, both quantitatively and qualitatively. The results of the uniaxial tensile test are presented in Fig. 4. The test is carried out by suspending the tube in a vertical configuration and applying a series of incremental weights. The resulting elongation is measured relative to the initial length (), enabling the characterization of the material response under axial loading (). The experimental data are compared with analytical predictions obtained from the application of the neo-Hookean model for this type of test (Eq. 5, from22). It can be observed that the model provides a satisfactory agreement with the experimental results when adopting shear moduli (G) of 0.37 MPa and 0.65 MPa for latex and rubber tubes, respectively. Normalized length () as a function of axial stress () for experimental data from uniaxial tensile tests on latex (G = 0.37 MPa) and rubber (G = 0.65 MPa) tubes, along with their corresponding analytical solutions (Eq. 5). The uniaxial results indicate that the neo-Hookean model is suitable for describing these materials, at least under this type of loading condition. The same assessment is next carried out for the test considered in this work, in which the tube is subjected to positive transmural pressure. Figure 5-left presents a comparison between experimental and numerical results for latex and rubber flexible tubes described in Table 1. The same shear moduli obtained from the uniaxial tests are used in the simulations. It can be observed that the model accurately predicts the response of both materials in both the near-linear regime (approximately below Pa for latex and Pa for rubber) and the nonlinear regime. The estimated shear moduli fall within the expected order of magnitude for both materials. Furthermore, these results are also compared with the Lamé solution for thick-walled cylinders subjected to transmural pressure, where Δ d and ν are, respectively, the increase in diameter and the Poisson coefficient (ν = 0.5 for both latex and rubber). These analytical results match with experimental data for both materials in the whole near-linear regime. However, in the non-linear regime the analytical model underestimates the volume. (Left) Inner volume as a function of transmural pressure according to experimental data, numerical simulations and Lamé theory for latex (G=0.37 MPa) and rubber (G=0.65 MPa). ν = 0.5 for both materials; (Right) Comparison of the external shape between simulations and experiments of the latex tube. Flexible tube dimensions are detailed in Table 1. The final step in the validation of the numerical model consists of a qualitative comparison between the shapes of the external surface of the flexible tube obtained from the experiments and those predicted by the simulations. Figure 5-right shows, for the latex tube, pictures of the experiment and the simulated shape overlapped at different volumes and transmural pressures. It is observed that the simulation matches the external shape of the flexible tube, which, together with the previous results, demonstrates the suitability of the Neo-Hookean model. These results also show that, in these experiments under positive transmural pressure, there is a nearly uniform diameter along most of the tube length, while the change in diameter occurs mainly in the regions close to the rigid tubes. These regions increase with transmural pressure. The incompressible neo-Hookean model employed in this study constitutes one of the simplest hyperelastic constitutive formulations, which inherently restricts its applicability to specific materials and loading regimes. However, it is well established in the literature that this model provides a reasonable approximation for elastomeric materials such as rubber, soft silicones (e.g., PDMS), and unfilled hydrogels, particularly under moderate deformation levels22,48. Its validity is generally limited to small to moderate strains, while more advanced models, such as the Ogden or Mooney–Rivlin formulations, are known to more accurately capture the nonlinear response observed at large deformations16,49,50. These material-related limitations do not significantly affect the objectives of this study, which focuses on the characterization of flexible elastomeric materials. The deformation levels considered here remain within a moderate range (Fig. 5-right), under which the neo-Hookean model is known to provide a consistent approximation of the mechanical response. Furthermore, the model predictions show good agreement with the experimental data, supporting its suitability for the present analysis. Nevertheless, it should be noted that the proposed approach is not intended to capture highly nonlinear behavior at large strains, where more advanced constitutive models may be required. Within these bounds, the numerical method is considered sufficiently validated and is subsequently employed to perform a parametric study aimed at characterizing the shear modulus of flexible tubes with varying thicknesses and lengths. Dimensionless parametric study As explained in section 2.1, the parameters that determine the deformation of the flexible tube (characterised as volume V) are the dimensions d, l and e, the transmural pressure Δ p, and the shear modulus G. This section performs a dimensionless study of the influence of these parameters (the Caret symbol is used to denote the dimensionless parameters). Pressure is non-dimensionalised with the shear modulus; volume is non-dimensionalised with the rest volume, ; and the length and thickness are non-dimensionalised with the rest radius, d/2. Then, the dimensionless deformation is reduced to the following expression, where , and are, respectively, the transmural load, the dimensionless thickness and the dimensionless length. Influence of thickness Figure 6-left shows a parametric sweep of the deformation as a function of the transmural load for different tube thicknesses and the same length (). As the flexible tube becomes thinner, a lower transmural load is needed to achieve the same deformation. It is observed that for values of between and 1 an increment of one order of magnitude in the thickness corresponds also to an increment of one order of magnitude of the transmural load. This effect is no longer observed when the wall thickness of the tube is relatively high (). (Left) as a function of for different and constant ; (Right) Deformed inner radius () along the tube axis for different with constant and . To further explore these results, Fig. 6-right presents a comparison of the deformed shapes of the flexible tube for different wall thicknesses at constant volume and length ( and ). For relatively thin-walled tubes ( and ), results show convergence in the shape of the inner wall. Also, is almost uniform along the tube except for the regions close to both ends. However, for relatively thick-walled tubes ( and ) this is not observed. Influence of length Figure 7-left shows a parametric sweep of the deformation as a function of the transmural load for different tube lengths and the same thickness (). As the flexible tube becomes longer, the transmural load needed to achieve the same deformation is lower, until the results converge for higher than 10. For relatively short tubes ( and ), there is no uniform region (Fig. 7-right). In addition, the axial gradient becomes steeper and the maximum increases as the length decreases. In contrast, when the simulated deformation close to the ends of the tube converges, and a uniform radius is achieved along most of the tube. These are the cases where becomes independent of . (Left) as a function of for different and constant ; (Right) Deformed inner radius () along the tube axis for different with constant and . Influence of transmural load-to-thickness ratio These observed phenomena show that, under certain conditions to achieve a given deformation, the required transmural load is proportional to the thickness of the wall. This behavior can be exploited to collapse the results and represent the deformation as a function of a single parameter, the load-to-thickness ratio, . This is plotted in Fig. 8, where this collapse can be observed for . The resulting curve from this collapse can be used to characterise the mechanical behavior of an elastomeric tube within the neo-Hookean framework. Knowing the initial dimensions of the tube, the applied pressure, and the deformed volume, it is possible to obtain the shear modulus of a flexible hyperelastic tube. To facilitate this, a function describing this collapse behavior is provided: As shown in Fig. 8, the proposed function accurately reproduces the tube behavior for relatively thin tubes. In this regime, increasing the tube length does not affect the results. However, the function does not capture the behavior of thick tubes (), regardless of the simulated length. This has also been verified for larger lengths (e.g., ). Moreover, the model has not been validated for this class of elastomeric tubes. Therefore, it is concluded that Eq. 8 can be reliably applied only to sufficiently long and thin elastomeric tubes, for which the tube volume can be determined using the proposed formulation and subsequently used for material characterization. as a function of for different thicknesses and . Dashed lines show the results from Eq. 8. Conclusions This paper numerically and experimentally studies the deformation of elastomeric flexible tubes subjected to positive transmural pressure. The numerical study consisted of static axisymmetric simulations in which the solid used a neo-Hookean constitutive model with the shear modulus as a single parameter. The numerical model was validated by comparing its results with two experiments performed with different flexible materials. In order to simplify the analysis, a dimensionless study was conducted. In this study, dimensionless deformed volume was analyzed as a function of transmural load (the ratio of transmural pressure to shear modulus) for different dimensionless lengths and thicknesses. The results showed that below a certain thickness, the length of the tube does not affect the deformation. With regard to thickness, for relatively thin-walled tubes, it has been observed that a decrease of one order of magnitude in thickness requires a lower transmural load of approximately the same order to produce the same state of deformation. For long enough thin-walled flexible tubes (so its deformation is independent of the length), the deformation behavior collapses into a function, when the dimensionless deformed volume is represented versus the ratio of the transmural load to the dimensionless thickness. Thanks to this curve, the shear modulus of a hyperelastic flexible tube can be obtained through a single experiment, requiring only a pressure gauge as instrumentation. Author contributions M.R.: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Supervision, Project administration. J.A.: Writing – review & editing, Writing – original draft, Validation, Investigation, Methodology, Data curation. C.B.-C.: Writing – review & editing, Writing – original draft, Validation, Investigation, Formal analysis, Conceptualization. M.A.H.: Validation, Investigation, Formal analysis, Conceptualization, Software, Supervision. Funding The authors would like to thank the Junta de Castilla y León for funding this work as part of the program “Subvenciones del programa de apoyo a proyectos de investigación financiados por fondos FEDER”, project number VA182P20, as well as the Spanish Ministry of Science and Innovation, grant no. PID2022-14095OB-C21. Data availability The datasets used during the current study will be available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used during the current study will be available from the corresponding author on reasonable request.