Hemodialysis is a life-sustaining therapy for patients with end-stage renal disease, in which blood is circulated outside the body for filtration and then returned to the vascular system. A reliable vascular access is provided by an arteriovenous fistula (AVF), a surgically created connection between an artery and a vein. During each dialysis session, arterial and venous needles are inserted into the fistula to withdraw and return blood. While clinically routine, repeated needling significantly alters local hemodynamics and has been linked to access-related complications such as stenosis, thrombosis, and eventual fistula failure. These outcomes are believed to be strongly influenced by complex, unsteady flow and stress environments that are difficult to interrogate directly in vivo. The goal of this project is to develop a high-fidelity computational framework, based on Large Eddy Simulation (LES), to characterize the flow structures and stress distributions associated with venous needling during hemodialysis. At the current stage, the simulations focus on simplified and idealized geometries representing venous return alone, enabling systematic investigation of jet–crossflow interaction, secondary flow formation, and unsteady shear stresses under clinically relevant flow conditions. Ongoing work involves validation of the numerical results against well-controlled laboratory experiments using idealized venous needling configurations. This validation step is essential to establish confidence in the predictive capability of the LES model. Once validated, the model will be extended to more realistic and clinically relevant scenarios. These include combined arterial and venous needling, incorporation of full AVF geometries, and the use of patient-specific vascular reconstructions. The numerical framework will further allow inclusion of physiological features that are challenging to realize experimentally, such as the non-Newtonian rheology of blood, pulsatile flow conditions, and complex three-dimensional vessel curvature. High-resolution simulations are required to resolve transient flow structures and stress fluctuations that are hypothesized to play a critical role in endothelial dysfunction and intimal hyperplasia. Access to NAISS computational resources is crucial for performing these LES calculations at the spatial and temporal resolutions necessary for accuracy and stability. The resulting datasets will provide detailed insight into the fundamental mechanisms linking needling-induced flow disturbances to regions susceptible to stenosis during dialysis treatment. Beyond advancing scientific understanding, this work aims to deliver a computational tool that can support clinicians and researchers in identifying high-risk flow environments and improving cannulation strategies, ultimately contributing to longer-lasting and safer vascular access for hemodialysis patients.