High strain rate material models for epoxide and silicone adhesives

Key findings from this study

• The study found that a Modified Johnson-Cook material model, calibrated with experimental data, accurately predicts silicone adhesive behavior at high strain rates, achieving 76.2% predicted versus 70.4% experimental transmitted strain.
• The authors report that existing material models for epoxide adhesives, particularly Plastic-Kinematic and Johnson-Cook formulations, perform well for epoxide systems but require independent validation before application to silicone adhesives.
• The researchers demonstrate that adapting existing empirical models within commercial finite-element method software eliminates the need for new material model development while achieving acceptable predictive accuracy.

Overview

This work develops and calibrates a Modified Johnson-Cook material model for silicone adhesives under high strain rate loading. The approach enables modeling of silicone adhesive behavior without requiring new user-defined materials or additional commercial finite-element method software development. The authors also evaluate existing material models previously used for epoxide adhesives to assess their applicability to silicone systems.

Methods and approach

The study calibrated an empirical material model through experimental data fitting across multiple high strain rates. The Modified Johnson-Cook model was adapted for silicone adhesives and validated against split-Hopkinson pressure bar test data from ceramic/steel material couples bonded with silicone adhesive. Additional material models—including Plastic-Kinematic with Cowper-Symonds strain rate scaling and Johnson-Cook with Grüneisen equation of state—were tested numerically against equivalent epoxide bonded specimens to evaluate their predictive capability.

Results

The calibrated Modified Johnson-Cook model for silicone adhesives predicted 76.2% transmitted strain compared to experimental split-Hopkinson pressure bar values of 70.4%, demonstrating reasonable agreement for this material system. For epoxide adhesives, both the Plastic-Kinematic material model with Cowper-Symonds strain rate scaling and the Johnson-Cook model with Grüneisen equation of state produced accurate predictions. The Plastic-Kinematic approach predicted 95.5% transmitted strain while the Johnson-Cook variant predicted 97.6%, both closely matching the experimental value of 91.2%.

The Modified Johnson-Cook formulation proved sufficient to capture silicone adhesive response across the tested strain rate range. The implementation within existing commercial finite-element method software avoided the necessity for custom material model development. The work establishes that material models developed for epoxide systems require evaluation before application to silicone adhesives, as the two material classes exhibit different high strain rate behavior.

Implications

The availability of a calibrated material model for silicone adhesives enables more accurate numerical prediction of bonded joint behavior under impact and dynamic loading conditions. Implementation via the Modified Johnson-Cook framework reduces barriers to adoption, allowing researchers and engineers to incorporate silicone adhesive modeling into standard finite-element analyses without software customization. This capability extends the applicability of existing computational tools to adhesive systems previously lacking validated high strain rate models.

The differential performance of material models across epoxide and silicone systems indicates that adhesive material class significantly influences strain rate sensitivity. Future work should validate these models across broader ranges of adhesive formulations and loading conditions to establish the generalizability of the calibrations. The framework established here provides a pathway for systematic model development and validation for other adhesive systems.

Scope and limitations

This summary is based on the study abstract and available metadata. It does not include a full analysis of the complete paper, supplementary materials, or underlying datasets unless explicitly stated. Findings should be interpreted in the context of the original publication.