Reducing the impact of geometric errors in flow computations using velocity measurements

Area of Science:

• Biomedical Engineering
• Computational Fluid Dynamics
• Medical Imaging

Background:

• Numerical blood flow simulations rely on anatomical medical images and velocity measurements for calibration.
• The accuracy of computational geometry is often limited by the resolution of medical imaging, leading to potential errors.

Purpose of the Study:

• To investigate the impact of boundary conditions on the accuracy of blood flow simulations with uncertain geometry.
• To propose and evaluate a novel methodology using slip/transpiration boundary conditions to improve simulation accuracy.

Main Methods:

• Simulations were set up using anatomical medical images, with standard no-slip boundary conditions initially applied.
• Slip/transpiration boundary conditions were introduced and their parameters estimated using velocity measurements.
• Numerical experiments were conducted to compare results with standard boundary conditions.

Main Results:

• Inaccurate boundaries with no-slip conditions can lead to significant errors, particularly in pressure gradient calculations.
• The proposed method using slip/transpiration boundary conditions considerably improved the accuracy of estimated pressure gradients.
• The methodology also enhanced the accuracy of 3D velocity fields when vessel geometry was uncertain.

Conclusions:

• Slip/transpiration boundary conditions offer a robust approach to enhance blood flow simulation accuracy.
• This method is particularly beneficial in scenarios where vessel geometry derived from medical images is uncertain.
• The findings suggest a valuable improvement for computational fluid dynamics in cardiovascular research.