Advisor(s)
Jeffrey Holmes
Committee Member(s)
Johane Bracamonte
Roy Koomullil
School
School of Engineering
Document Type
Thesis
Department (new version)
Engineering
Date of Award
9-11-2025
Degree Name by School
Master of Science in Mechanical Engineering (MSME) School of Engineering
Abstract
Over one million Americans suffer from myocardial infarction (MI) per year. After MI, the heart muscle is replaced by scar, whose mechanical properties play a critical role in outcomes for surviving patients. Previously, Caggiano et al. induced MI in rats, and ex-tracted the scar tissue samples for microscopy and mechanical testing. Building on this, we investigate the effect of incorporating microscopy-derived fiber heterogeneity on the estimation of stress and material parameters. We studied the effect of fiber orientation heterogeneity on the mechanical characteriza-tion of scar tissues using finite element models (FEM) that replicated biaxial tests. We used three constitutive models—Mooney-Rivlin, Fung-Guccione, and Holtzapfel-Gasser-Ogden (HGO). We utilized data from four scar samples, including histology-derived fiber orientation by layers, and biaxial stretch-stress pairs. We explored three fiber orientation assumptions: (1) homogeneous and perfectly aligned, (2) homogeneous with histology-derived orientation and (3) heterogeneous, with histology-derived layer-specific orienta-tion. We analyzed mesh sensitivity, we studied the effect of fiber orientation on stress distribution, and estimated the set of material parameters that best reproduce the biaxial experimental data. The estimation of material parameters was carried through parameter screenings followed by the application of the Powell optimization algorithm. Optimiza-tion was applied to minimize the mean squared error (MSE) between simulated and ex-perimental mean stress over all the biaxial test protocols. Mesh sensitivity analysis showed heterogeneous models required at least 75 elements through the thickness to capture steep orientation gradients (~151°/cm). While heteroge-neity didn’t alter stress distribution patterns, it increased variability and extremes. Ex-treme values of stress were always found in the boundaries; however, estimated mean stress was unsensitive to boundary effects. We observed that multiple parameter sets pro-duced similar MSEs, and that low MSE could be achieved with linear-like solutions that misrepresent the data. This suggest that minimum MSE may not be a sufficient criteria for the identification of sample-specific optimum parameter sets. Furthermore, including histological alignment improved mechanical behavior prediction, though layer heteroge-neity didn’t always lower error. HGO parameters were consistent across models, suggest-ing microstructure-based models are robust and that simplified models can yield reasona-ble estimations.
ProQuest ID
Prior version
Recommended Citation
Flanagan, Desmond Maximilian, "Effect Of Fiber Structure Heterogeneity On The Estimation Of Material Parameters For Cardiac Scar Tissue Mechanics" (2025). All ETDs from UAB. 7374.
https://digitalcommons.library.uab.edu/etd-collection/7374
