Development of Finite Element Models for Characterizing Silicone Ring with Support Pessaries

Objectives: Mechanical tests, like pessary folding, can be used to characterize pessary properties but are resource intensive. Simulating the mechanical tests through finite element (FE) modeling can provide expedited results for pessary material and parameter analysis. This study aimed to identify viscoelastic material models for pessary silicones (Shore 60A and 40A) and create computational models of two pessary mechanical tests for ring with support (RWS) pessaries.

Methods: Silicone material characteristics were determined using scaled uniaxial tension tests on standardized “dog-bone” shaped samples (40A: n = 5, 60A: n = 6) and compression tests on cylinder samples (40A: n = 5, 60A: n = 5). The experimental data for each silicone was then fitted with a viscoelastic, Mooney-Rivlin (MR) material model by minimizing the difference between each model and its corresponding test data. RWS pessaries in standard sizes (3, 5, 7, 11) were manufactured in silicone using cocoon molding and subjected to two mechanical testing protocols. First, a folding test (replicating simplified pessary insertion mechanics) was performed by loading each pessary vertically, in its designed folding orientation, between a uniaxial load applicator and fixed plate until the pessary folds to self-contact (Figure 1A). Then, a modified 3-point bending test (replicating simplified pessary support) was performed by applying a load to each pessary horizontally suspended between two curved platforms until the pessary slipped from the platforms (Figure 1E). Subsequently, these tests were recreated in FE analysis software (FEBio v3.8, FEBio Studio v1.9). The previously identified MR material coefficients were applied to the size 5 FE model for each silicone, then iteratively refined to match the experimental folding test data. These optimized properties were then independently validated against FE simulations on sizes 3, 7, and 11 for the fold and bending tests. Agreement between the FE model and experimental test data was assessed using correlation and analysis software (CORAplus, v4.0.5).

Results: The MR material model coefficients were determined for both the Shore 60A (C1=0.6410, C2=-0.1238, K=51.3765) and 40A materials (C1=0.2981, C2=-0.0874, and K=20.9340). Test data and FE results for the calibrated (size 5) and independent models (sizes 3, 7, 11) were plotted (Figure 1). A strong agreement between the FE and mechanical test data was found with CORAplus evaluation (Table 1).

Conclusions: This study successfully developed viscoelastic materials models and computational models to evaluate RWS pessary folding and modified 3-point bending, as demonstrated by the strong agreement between the response of the FE models and mechanical tests. These results will allow computational testing and mechanical response evaluation to be performed on novel pessary designs to expedite the labour intensive iterative design process.