Modelling and prediction of cavitation erosion in GDi injectors operated with E100 fuel

FUEL

卷:289

Article

Santos, Eduardo Gomez^1,2  Shi, Junmei¹  Venkatasubramanian, Ramesh¹  Hoffmann, Guy¹  Gavaises, Manolis²  Bauer, Wolfgang³ 

[1] Delphi Technol, Ave Luxembourg, L-4940 Bascharage, Luxembourg

[2] City Univ London, Northampton Sq, London EC1V 0HB, England

[3] ANSYS Germany GmbH, Staudenfeldweg 20, D-83624 Otterfing, Germany

关键词: Cavitation;    Erosion;    E100 fuel;    Gasoline direct injection;    LES;    URANS;   

DOI  :  10.1016/j.fuel.2020.119923

来源: Elsevier

【 摘 要 】

Ethanol (E100) can be utilised in spark ignition engines for passenger car vehicles. This brings a challenge to the durability of the fuel injection system components since its use can result in corrosion, further enhanced by cavitation-induced erosion. This work reports computational fluid dynamics (CFD) predictions for both the flow development and the locations prone to cavitation erosion in multi-hole gasoline direct injection (GDi) injectors operated with E100. The compressible form of the Navier-Stokes equations is solved numerically considering the motion of the injector's needle valve. Thermodynamic and mechanical equilibrium is assumed between the liquid, vapour and non-condensable gas; E100 liquid and vapour are considered as a barotropic fluids where the corresponding variation in density with pressure and the speed of sound are estimated via a relevant equation of state; an additional transport equation is solved for simulating the non-condensable air entrainment into the injector during the dwell time between successive injections. Turbulence is modelled using both large eddy simulation (LES) and Unsteady Reynolds-averaged Navier-Stokes (URANS) considering a sector and the full nozzle geometry, respectively. Various cavitation erosion indices reported in the literature are evaluated against new durability tests of surface erosion damage obtained after 400 M injection cycles. The relevant nozzle wall erosion images are found to correlate well with the accumulated erosive power predicted from the computational model.

【 授权许可】

Free