A genetic algorithm optimization framework for the characterization of hyper-viscoelastic materials: application to...

Area of Science:

• Biomechanics
• Materials Science
• Computational Modeling

Background:

• Human articular cartilage (AC) exhibits complex hyper-elastic and viscoelastic properties.
• Accurate characterization of AC is crucial for understanding joint health and disease.
• Existing methods for material characterization can be time-consuming and complex.

Purpose of the Study:

• To develop an automated computational framework for characterizing hyper-elastic and hyper-viscoelastic materials.
• To apply and validate this framework using human articular cartilage (AC) samples.
• To optimize the material approximation using Prony series and computational algorithms.

Main Methods:

• Dynamic mechanical analysis (DMA) with frequency sweeps (1-90 Hz) on 26 human AC samples.
• Automated finite element analysis (FEA) integrated within a modular framework.
• Optimization of hyper-viscoelastic material models using a genetic algorithm and interior point technique.
• Evaluation of Prony series approximations (N=1, 3, 5) over 20 and 50 generations.

Main Results:

• The automated framework effectively characterized human articular cartilage.
• Approximations with N=3 and N=5 Prony series orders showed approximately 5% less error compared to N=1.
• A 10% error difference was observed during unloading between N=5 and N=1 approximations.
• Increasing genetic algorithm generations from 20 to 50 reduced parameter error by approximately 1%.

Conclusions:

• The developed automated framework is effective for characterizing human articular cartilage.
• The study demonstrates the utility of computational approaches for complex material modeling.
• The findings provide a foundation for advanced biomechanical simulations and analyses of articular cartilage.