Abstract:The personalisation of finite element models is an important problem in the biomechanical fields where subject-specific analyses are fundamental, particularly in studying soft tissue mechanics. The personalisation includes the choice of the constitutive law of the model's material, as well as the choice of the material parameters. In vivo identification of the material properties of soft tissues is challenging considering the complex behaviour of soft tissues that are, among other things, non-linear hyperelastic and heterogeneous. Hybrid experimental-numerical methods combining in vivo indentations and inverse finite element analyses are common to identify these material parameters. Yet, the uniqueness and the uncertainty of the multi-material hyperelastic model have not been evaluated. This study presents a sensitivity analysis performed on synthetic indentation data to investigate the identification uncertainties of the material parameters in a bi-material thigh phantom. Synthetic numerical data, used to replace experimental measurements, considered several measurement modalities: indenter force and displacement, stereo-camera 3D digital image correlation of the indented surface, and ultrasound B-mode images. A finite element model of the indentation was designed with either Ogden-Moerman or Mooney-Rivlin constitutive laws for both materials. The parameters' identifiability (i.e. the possibility of converging to a unique parameter set within an acceptable margin of error) was assessed with various cost functions formulated using the different synthetic data sets. The results underline the need for multiple experimental modalities to reduce the uncertainty of the identified parameters. Additionally, the experimental error can impede the identification of a unique parameter set, and the cost function depends on the constitutive law. This study highlights the need for sensitivity analyses before the design of the experimental protocol. Such studies can also be used to define the acceptable range of errors in the experimental measurement. Eventually, the impact of the evaluated uncertainty of the identified parameters should be further investigated according to the purpose of the finite element modelling.

Identification of non-linear behavior models with restricted or redundant data

Viscoelastic Bilinear Cohesive Model and Parameter Identification for Failure Analysis of Adhesive Composite Joints

An Efficient Multi-parameter Synchronous Identification Method for Fiber-Reinforced Laminated Structure Based on Improved Levenberg–Marquardt Algorithm and Modal Data

Identification of material parameters in low-data limit: application to gradient-enhanced continua

Data-Driven Accelerated Parameter Identification for Chaboche-Type Visco-Plastic Material Models to Describe the Relaxation Behavior of Copper Alloys

Automatic parameter identification of a shape memory alloy model using characteristic experimental data points

Nonlinear material identification of heterogeneous isogeometric Kirchhoff-Love shells

Cyclic behavior of laminated bio-based connections with slotted-in steel plates: Genetic algorithm, deterministic neural network-based model parameter identification, and uncertainty quantification

Statistical analysis of identification of linear viscoelastic models

Identifiability of material parameters in solid mechanics

Constitutive Modelling of Laminated Fibre-Reinforced Composites

Hybrid Inverse Parameter Identification of Fully Coupled Ductile Damage Model for Steel Sheet DP600 with Two Different Algorithms: Trust Region and Genetic Algorithms

Efficient parameters identification of a modified GTN model of ductile fracture using machine learning

Identification of constitutive materials of bi-layer soft tissues from multimodal indentations

Identification and characterization of damaged fiber-reinforced laminates in a Bayesian framework

Automated identification of linear viscoelastic constitutive laws with EUCLID

Quantification of the out-of-plane loading fatigue response of bistable CFRP laminates using a machine learning approach

Model-free data-driven methods in mechanics: material data identification and solvers

Bayesian Approach to Micromechanical Parameter Identification Using Integrated Digital Image Correlation

Damage identification in fiber metal laminates using Bayesian analysis with model order reduction

Machine learning meta-models for fast parameter identification of the lattice discrete particle model