Browsing by Author "Monokrousos, Nikos"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    A numerical approach to overcome the very-low Reynolds number limitation of the artificial compressibility for incompressible flows
    (Elsevier, 2024-11-15) Monokrousos, Nikos; Könözsy, László Z.; Pachidis, Vassilios; Sozio, Ernesto; Rossi, Federico
    We propose a numerical approach to solve a long-standing challenge which is the applicability of the artificial compressibility (AC) formulation for solving the incompressible Navier—Stokes equations at very-low Reynolds numbers. A wide range of engineering applications involves very-low Reynolds number flows in Micro-ElectroMechanical Systems (MEMS) and in the fields of chemical-, agricultural- and biomedical engineering. It is known that the already existing numerical methods using the AC approach fail to provide physically correct results at very-low Reynolds numbers (Re ≤ 1). To overcome the limitation of the AC method for these engineering applications, we propose a higher-order Neumann-type pressure outflow boundary condition treatment along with their up to fourth-order numerical approximations. We found that the numerical treatment of the pressure at the outlet boundary plays the main role in overcoming the limitation of the AC method at very-low Reynolds numbers (Re << 1). Therefore, we provide numerical evidence on the accuracy of the AC method beyond its previously reported limitations, e.g., the low Reynolds number Oseen flow (Re << 1) is first presented in this work. A third-order explicit total-variation diminishing (TVD) Runge–Kutta scheme has been employed with standard finite difference spatial discretisation schemes for improving the accuracy of the numerical solution. For modelling strongly viscous flows, the Reynolds number ranges from 10-1 to 10-4. Overall, we found that the accuracy limitation of the AC method below Re < 1 can be overcome with an accurate numerical treatment of the outlet pressure boundary condition instead of using high-order schemes in the governing equations. For the investigated Reynolds number range (10-1 ≤ Re ≤ 10-4 ), the obtained results show that the relative errors were smaller than 1% for the numerical simulations performed on the configurations of both the two- and three-dimensional, straight microfluidic channels. The imposition of high-order derivative Neumann-type pressure outflow boundary conditions reduced the maximum relative errors of the numerical solutions from 85% and 95% to below than 1% at the outlet section of the two- and three-dimensional, straight microfluidic channel flows, respectively. Taking the advantage of the numerical approach proposed here, two- and three-dimensional benchmark problems employed in the current investigation in comparison with analytical solutions available in the literature, clearly demonstrate that the artificial compressibility can be used beyond its previously known constraints for very-low Reynolds number incompressible flows.
  • Thumbnail Image
    ItemOpen Access
    Numerical solution of LOx flow in a liquid rocket engine additively manufactured cooling channel
    (International Council of the Aeronautical Sciences, 2024-10-08) Monokrousos, Nikos; Könözsy, László Z. ; Pachidis, Vassilios; Sozio, Ernesto; Rossi, Federico
    The present work has been conducted in the framework of the DemoP1 demonstrator for the design of a LOx/LNG aerospike engine carried out by Pangea Aerospace. The main objective of the demonstrator is to highlight the central features of the next generation booster-class engines such as reusability, the utilisation of cryogenic coolants and the benefits arising from the ever-growing field of Additive Manufacturing (AM) for high heat flux aerospace applications with increasing thermal load management demands. In the present study a numerical investigation of the cryogenic liquid oxygen coolant flow in an AM cooling channel of the DemoP1 engine is implemented. The simulations are performed on a three-dimensional curvilinear cooling channel of variable, rectangular cross-section of the aerospike engine. Different variations of the two-equation k − ω turbulence model are employed and assessed for the closure of the fluid flow governing equations and the identification of the efficient formulations for the accurate prediction of the spatial development of the primitive variables. The numerical solutions obtained for the characterisation of heat transfer and pressure drop in the AM cooling channel are compared against experimental data provided from Pangea Aerospace for the full-scale single-injector element hot-fire test campaign of the DemoP1 aerospike engine demonstrator.