PSI - Issue 45

Xiaochen Wang et al. / Procedia Structural Integrity 45 (2023) 88–95 Xiaochen Wang/ Structural Integrity Procedia 00 (2023) 000 – 000

91

4



  1 e



1 d

c

3

3

 2 3) 1     (2) where a i , b i , c 1 , c 2 , e 1 , and e 2 are non-dimensional constant parameters. ILT was assumed to be isotropic, incompressible and hyperelastic, which was modelled using a two parameter Mooney-Rivlin model (Li et al., 2008). The blood is considered as a non-Newtonian and incompressible fluid to include the shear-thinning effect. This study utilises nonlinear models of large deformation through FEM-based FSI. Four models were created: two with isotropic material models and two with anisotropic material models. In each group, there was also a model that included the presence of ILT on the inner wall of the aorta. The idealised geometries used in these models are depicted in Fig. 1 and were based on primary dimensions obtained from CT images of a patient from the Royal Adelaide Hospital. The calculations were performed using tetrahedral elements and the grid independency was assessed. The governing equations for fluid, AAA, and ILT domains are described in Eqn. (3) to (5), respectively (Throop et al., 2022, Bantwal et al., 2021). These equations characterisse the blood and AAA density  f and  A , velocity fields of the fluid, mechanical and AAA domains, U f , U c and U A , and the Cauchy stress tensor for fluid, AAA and ILT  f ,  A and  ILT . [ ] 0, 0, in fluid domain , f f f c f f f f t                  U U U U U (3) 0, in AAA domain , A A A A A A g t                U U U (4)   0, in ILT domain . ILT ILT     (5) The models in this study were subjected to pulsatile parabolic velocity profiles at the inlet and a pulsatile pressure at the outlet, with the inflow and outflow boundary conditions applied in the proximal and distal regions, respectively (Scotti and Finol, 2007). The wall was assumed to be a non-slip surface and fixed supports were imposed at both the inlet and outlet. The rest of the geometry was allowed to deform in all directions. The models were solved through ANSYS (version 2020R2, ANSYS, Inc., Cannonsburg, PA) with spatial discretisation of pressure and a maximum of 100 iterations per substep. To mitigate the effects of the initial flow conditions, the model was pre-run with two cardiac cycles. (a) (b) 2 ( 3) 1 ,    1 1   2 2 4 2 6 exp e I  1 1 2 2 1 ( 3) ( b I 3)   ( exp c I  2 2 i   j i i i j J a I c e            2

Fig. 1. (a) CT scan image from patients in axial plane; (b) idealised geometries for models without and with ILT.

3. Results and discussion In this section, the results of the numerical simulation of the models include wall stress mapping, wall deformation patterns, wsdistributions and blood flow fields. These results are compared in order to investigate the effects of ILT and material models on the AAA wall rupture mechanics.