Abdominal aortic aneurysms (AAAs) are focal dilatations of the infrarenal aortic wall, often caused by localized weakening of the vessel wall. While most AAAs are asymptomatic, progressive expansion can lead to rupture with severe, life-threatening medical consequences. Untreated, a ruptured aneurysm carries a mortality rate as high as 85%, and even with timely medical intervention, the mortality rate exceeds 30% (Kent, 2014). This underscores the critical importance of accurately assessing AAA rupture risk. Clinically, the maximum diameter is used as an indicator for rupture potential and surgical intervention. This criterion is known to underestimate the rupture risk of small AAAs, especially in female patients (Moll et al., 2011). Additionally, for some patients with large AAAs, the risks associated with surgery may outweigh the risk of aneurysm rupture (Lederle et al., 2002). Consequently, a comprehensive risk assessment of AAAs based on multiple factors is essential for optimized patient management.

The arterial wall consists of three distinct layers: the intima, media, and adventitia, each playing a critical role in maintaining vascular function and regulating hemodynamic forces (Holzapfel et al., 2000). The intima, composed primarily of endothelial cells, provides a smooth surface for blood flow; the media, rich in smooth muscle cells, elastin, and collagen, modulates vascular diameter and elasticity, enabling the vessel to endure arterial pressure; the adventitia, composed of loose connective tissue, provides tensile strength and flexibility, thereby maintaining the structural integrity of vessels. Vascular smooth muscle cells within the media play a crucial role in processing the extracellular matrix, including the formation of elastin and collagen fibers. The degradation of these fibers and the reduction of the smooth muscle cells significantly contribute to the progression of AAAs.

To describe the material properties of arterial walls, Holzapfel et al. proposed using an isotropic model to represent the non-fibrous matrix and employed an exponential function to describe the fiber behavior (Holzapfel et al., 2000). Building on that, Gasser et al. introduced a fiber dispersion factor, capturing the symmetric fiber distribution around a mean orientation (Gasser et al., 2006). Holzapfel et al. later refined this model by incorporating a non-symmetric orientation density function, offering a more detailed representation of fiber distributions (Holzapfel et al., 2015). Several experimental studies have applied these models, or their variants, to determine layer-specific material parameters for both healthy and aneurysmal vessel walls (Weisbecker et al., 2012; Sassani et al., 2015; Niestrawska et al., 2016).

Intraluminal thrombus (ILT) refers to blood clots formed within vessels. Approximately 75% of AAAs contain thrombi (Tong and Holzapfel, 2015). While ILT typically functions to stop bleeding and repair blood vessels as part of the normal physiological process, it complicates the rupture risk assessment. Some studies suggest that aneurysm rupture may be correlated with increased ILT volume (Hans et al., 2005; Haller et al., 2018). A thicker ILT may lead to hypoxia in the adjacent AAA tissue, which may result in the degradation of the extracellular matrix and a significant reduction of wall strength (Vorp et al., 2001). Additionally, ILT may promote vascular smooth muscle cell apoptosis, leading to a thinner vascular wall beneath it (Koole et al., 2013). As a result, ILT can contribute significantly to AAA growth and elevate the rupture risk. Conversely, from a biomechanical standpoint, the thrombi also reduce the stress on the underlying AAA tissue, which may lower the rupture risk (Wang et al., 2002; Di Martino and Vorp, 2003; Throop et al., 2022). As a result, when assessing the rupture risk of AAA, the impact of ILT needs to be thoroughly studied.

Finite element analysis and fluid-structure interaction (FSI) are valuable tools for assessing the rupture risks of AAAs and understanding the biomechanical impact of ILT by evaluating the wall stress of AAAs. Traditional finite element analysis employs an idealized geometry and applies uniform internal pressure to simulate the load exerted by blood flows. Although this method offers computational simplicity, it fails to account for the pulsatile nature of blood flow, a critical factor in both AAA rupture and ILT formation. Consequently, this simplification may lead to an incomplete assessment of rupture risk. FSI offers more realistic simulation results by integrating the interaction between fluid (e.g., blood) and structure (e.g., arterial wall and ILT). It is noted that the enhanced accuracy of FSI comes at the cost of increased modeling complexities. Early studies relied on idealized geometric models and isotropic material assumptions for both the aneurysm and ILT (Di Martino and Vorp, 2003; Scotti et al., 2005). Although these models provided valuable insights, they lacked the fidelity to accurately represent the patient-specific geometries and physiologically realistic material properties.

With advancements in medical imaging, patient-specific geometries of AAAs and ILTs have been incorporated into both finite element and FSI analyses, enabling more accurate predictions of wall stress distribution (Maier et al., 2010). This enhanced modeling capability is crucial for assessing rupture risk, particularly when combined with the spatial distribution of wall strength. Given the anisotropic nature of vascular tissues, contemporary analyses have increasingly adopted anisotropic material models (Rodríguez et al., 2009; Riveros et al., 2015; Rissland et al., 2009; Xenos et al., 2010; Xenos et al., 2015). Studies have demonstrated that stress magnitudes derived from anisotropic models are significantly higher than those from isotropic models, underscoring the importance of incorporating anisotropic models to accurately represent material properties (Rodríguez et al., 2009; Rissland et al., 2009). Moreover, accounting for the layered tissue structure allows for a more refined transmural stress distribution, as each layer exhibits distinct mechanical behaviors essential for a comprehensive understanding of AAA biomechanics (Simsek and Kwon, 2015; de Lucio et al., 2021). Notably, degeneration of the media has been incorporated to explore mechanisms underlying AAA initiation and progression (Simsek and Kwon, 2015). Regarding the impact of ILT on AAA stress conditions, most studies indicated that ILT significantly affects the stress distribution on the aneurysm wall, with the maximum wall stress typically occurring in regions where the ILT is thinnest (Riveros et al., 2015). Interestingly, one study noted that while ILT affects stress magnitude, it does not significantly alter the site of the peak stress (Xenos et al., 2015).

Given these inconsistent findings, it is evident that current models may not fully capture the complexities of AAA biomechanics. Our research aims to address these gaps by developing a more comprehensive approach that incorporates the layered structure of the vascular wall, anisotropic material properties, ILT, and ILT-induced media degradation. To achieve these objectives, this work is organized as follows. Section 2.1 presents the geometric modeling method for ILT and the layered architecture of the AAA wall. It also details the procedure for generating a local coordinate system for the solid mesh, which facilitates the description of the anisotropic tissue models. Section 2.2 presents the FSI formulations employed in this study, including the physiological boundary conditions for the simulations. Section 2.3 introduces the material models applied in the simulations, including the anisotropic hyperelastic model for the AAA tissue and the isotropic hyperelastic model for the ILT. Section 3 presents the FSI analyses conducted on both fusiform and saccular AAAs. For cases with multi-layered tissue model, layer-specific material parameters are used to describe the intima, media, and adventitia of the AAA wall. We compare the maximum principal stress (MPS) distribution, focusing on layer-specific stress distributions under three conditions: (1) without ILT, (2) with ILT but no tissue degradation, and (3) with both ILT and degradation. We discuss our findings in Section 4 and draw conclusions in Section 5. Limitations of this study is discussed in Section 6.

In this section, we perform FSI analysis and discuss the results. Section 3.1 compares the results obtained with single-layered versus multi-layered wall models; Section 3.2 examines the influence of ILT using multi-layered wall model by comparing cases with and without ILT; in Section 3.3, we investigate the media degradation caused by ILT by virtually introducing a degradation zone in the media. In all three sections, the average MPS is calculated to capture the overall stress distribution in the region of interest, enabling comparisons across cases. In Sections 3.1, 3.2, we also identify the location of the maximum MPS, which is essential for determining the regions within the highest rupture risk. Comparing these high-stress locations across cases, we gain insights into the influence of different factors on wall stress distribution.

In this study, the fluid density is set to 1.0 g ⋅ c m − 3 and its dynamic viscosity is set to 0.04 p o i s e . The material parameters of the tissue and ILT are listed in Table 1. We mention that the material parameters of the ILT are adopted from the prior studies (Di Martino and Vorp, 2003), and the parameters of the tissue layers are given by layer-specific uniaxial tensile tests of AAA tissues (Sassani et al., 2015). The tests of (Sassani et al., 2015) were conducted on samples from patients undergoing open repair of AAAs. Each layer underwent uniaxial tensile tests in both axial and circumferential directions under physiological conditions. Material parameters were obtained using a nonlinear least-squares fitting procedure based on the material model presented in (Gasser et al., 2006). The fitting process incorporated both circumferential and axial data simultaneously to ensure consistency. The parameters for the single-layered wall model are obtained by averaging the parameters of the three layers, with weights based on their thicknesses. For the degraded media, the parameters μ and k 1 are set to 5% of their original values.

To gain insights into the above material models and parameters, we consider a thin-walled tube problem as a preliminary analytical test of the anisotropic material model (Gasser et al., 2006). Assuming incompressibility, the internal pressure p in , circumferential stretch λ c , and axial stretch λ a are related by the following equations, λ a ∂ G ̂ d ∂ λ a − λ a λ c R − H / 2 λ a λ c 2 2 H R p in = 0 , (5) λ c ∂ G ̂ d ∂ λ c − λ c 2 λ a R H − 1 2 p in = 0 , (6) where R represents the mean radius of the tube, and H denotes the wall thickness. We set R = 10.0 mm and H = 0.9 mm to mimic a short segment of the infrarenal aorta. The modified free energy G ̂ d is derived based on the model (Equation 4) and adopts the form, G ̂ d λ a , λ c = μ 2 I ̂ 1 − 3 + k 1 k 2 exp k 2 κ I ̂ 1 + 1 − 3 κ I ̂ 4 − 1 2 − 1 , where the invariants I ̂ 1 and I ̂ 4 are given as, I ̂ 1 = λ a 2 + λ c 2 + 1 / λ a λ c 2 , I ̂ 4 = λ a 2 c o s 2 θ + λ c 2 s i n 2 θ . Solving (Equations 5, 6) with a series of prescribed values of p in , we obtain a plot of p in versus λ c , as shown in Figure 8. Among the three layers, the media exhibits the strongest mechanical response to the loads under the same circumferential stretch. This is primarily due to the fact that the mean orientation of the media is more aligned to the circumferential direction than in the other two layers, and the fibers are less dispersed in the media.

In the FSI analysis, the inlet flow boundary condition is depicted in Figure 9. The Vascular Model Repository provides the original inlet flow boundary condition at the proximal end of the descending aorta (VMR, 2024). The original flow rate is scaled based on the blood flow distribution ratio to determine the inlet flow rate of the AAAs. In this study, the distribution ratio is taken as 28.2% for both AAA models. The parameters for the outflow boundary condition are listed in Table 2. The parameters of the lumped parameter model are determined through parameter tuning, ensuring that the outlet pressure remains within the physiological range. The outlet pressure curves, shown in Figure 10, follow a similar pattern across all cases in this study, with the pressure value ranging from 70 mmHg (diastolic pressure) to 105 mmHg (systolic pressure), reflecting typical physiological conditions.

For the fAAA (sAAA) cases, the time step size is set to be 5.0 × 1 0 − 5 s ( 6.8 × 1 0 − 5 s ) . All displayed results are obtained at the moment when the tissue reaches its peak stress during the second cardiac cycle. Mesh independence has been confirmed, and we present the results from the finest mesh. The mesh details are listed in Table 3.

Previous studies have applied layered models for stress analysis in cardiovascular diseases, and they reported similar stress distributions across the tissue layers, with the media experiencing the highest stress, followed by the adventitia, and then the intima (Gao et al., 2013; Simsek and Kwon, 2015; Fan et al., 2024). This suggests that the media serves as the primary load-bearing layer of the vessel under physiological loading. Likewise, the work of Simsek suggests that when the media degrades, the intima and adventitia take on a greater load (Simsek and Kwon, 2015), aligning well with our findings. According to Sassani, in the case of three-layered wall rupture, the stress in the adventitia exceeds that in the media, with both layers bearing higher stress than the intima. This highlights the adventitia’s role in maintaining vascular structural integrity when the media degrades (Sassani et al., 2015). From this discussion, it is clear that compared to single-layered tissue models, multi-layered tissue models provide valuable insights into the biomechanical responses of the individual layers and their distinct roles in bearing physiological loads. Therefore, when analyzing stress distributions in vascular tissues, a physiologically detailed model should be considered. However, we also note that there is no significant difference in the time-averaged wall shear stress between the single-layered and multi-layered models, indicating that hemodynamic forces are less sensitive to the choice of the tissue model. This is consistent with our prior study (Sun et al., 2025a).

Numerous studies have investigated the biomechanical effects of ILT, especially regarding its impact on the location of the maximum stress. Throop et al. emphasized that it can significantly alter the location of the peak stress (Throop et al., 2022). Riveros et al. found that the region of the maximum stress often coincides with the thinnest part of the ILT (Riveros et al., 2015). Conversely, Xenos et al. reported that ILT had little impact on the location of the maximum stress (Xenos et al., 2015). Our findings suggest that whether ILT affects the location of the maximum MPS depends on whether the ILT sufficiently covers the area of the peak stress. These varying results underscore the complex role ILT plays in aneurysm biomechanics and highlight the need for further patient-specific studies to better understand its effects on stress distribution and rupture risk.

Our study demonstrates that, under equivalent conditions of degradation, a thinner ILT provides significantly less biomechanical protection to the affected region compared to a thicker ILT. However, in the specific cases analyzed in this study, even with increased stress levels observed in the intima and adventitia after degradation, the stress levels remained lower compared to the failure stress reported by Sassani et al. (2015), which are 276.8 kPa (axial) and 511.1 kPa (circumferential) for intima, 957.5 kPa (axial) and 1728.2 kPa (circumferential) for adventitia, even when the ILT was relatively thin. This suggests that, in these scenarios, the presence of a thinner ILT does not substantially increase the rupture risk. Since our model constructed the degradation regions virtually, we believe that more sophisticated models are needed to explore the impact ILT might have on aneurysm rupture. This should include accounting for the local thickness of the ILT (Vorp et al., 2001) and the age of ILT formation (Tong et al., 2011). These factors are crucial for accurately predicting the protective or detrimental effects of ILT on aneurysm stability.

In this study, we investigated the biomechanical behavior of AAA by incorporating the layered architecture of the vascular wall, anisotropic material properties, and the effects of ILT, particularly its role in media degradation. Through detailed FSI analysis of both fusiform and saccular AAAs, we compared the MPS distribution under various conditions: with ILT, with ILT but no degradation, and with both ILT and degradation. The results offer valuable insights into the stress variations within each layer of the aneurysm wall, enhancing our understanding of AAA biomechanics and the potential impact of ILT on aneurysm progression.

The multi-layered AAA tissue model, compared to the single-layer model, offers a more detailed transmural stress distribution, with the media serving as the primary load-bearing component of the aneurysm tissue. Moreover, the presence of ILT significantly reduces the stress levels in the aneurysm wall beneath it. However, ILT does not necessarily affect the location of the maximum stress. Degradation of the media increases stress levels in both the intima and adventitia.

In the future, we will build upon the current multi-layered anisotropic hyperelastic model by incorporating a viscoelastic model to better capture the biomechanical properties of vascular tissues. By combining this approach with existing FSI tools, we will be able to more accurately model cardiovascular and cerebrovascular diseases, thereby enhancing our understanding of the biomechanical mechanisms underlying these conditions.

The sample size used in the experiments is relatively small. While the data collected provide valuable insights into the studied topic, the limited sample size may affect the generalizability of the findings to a broader population. Specifically, the small sample size introduces a risk of bias due to individual differences. For instance, AAAs with a maximum diameter exceeding 5.5 mm were not included in this study, which may limit the applicability of our results to small AAAs. Nevertheless, the main conclusions of this study remain consistent with findings from existing studies (Simsek and Kwon, 2015; Xenos et al., 2015; Riveros et al., 2015).

Furthermore, for the purpose of using FSI to diagnose AAA, obtaining additional patient-specific information, such as material properties, is essential and challenging. In this study, we used the same material parameters derived from experimental data to describe ILT and AAA tissue (Di Martino and Vorp, 2003; Sassani et al., 2015). However, as the results obtained using the multi-layered model align with both the functional roles of different layers (Holzapfel et al., 2000) and the experimental findings (Sassani et al., 2015), the conclusions of this study remain reliable.