Well-balanced high-order method for non-conservative hyperbolic PDEs with source terms: application to one-dimensional blood flow equations with gravity

The present work proposes a well-balanced finite volume-type numerical method for the solution of non-conservative hyperbolic partial differential equations (PDEs) with source terms. The method is characterized, first, by the use of a recently introduced high-order spatial reconstruction, based on generalized Riemann problem information from the previous time level. Such reconstruction is well-balanced up to order three, compact, efficient and easy to implement. Second, the method incorporates a well-balanced space-time evolution operator, which allows for well-balanced fully explicit time evolution. The accuracy and efficiency of the method are assessed on both a scalar problem (Burgers' equation) and a nonlinear PDE system (hyperbolized one-dimensional blood flow equations with gravity and friction, and with variable mechanical and geometrical properties). The well-balanced property is verified by showing that numerically-determined stationary solutions are preserved up to machine precision. The order of accuracy in space and time is validated through empirical convergence rate studies. Additionally, the performance of the method is assessed on a network of 86 arteries, under both stationary and transient conditions.