Final Report Summary - CAVFUELSYSTEM (Cavitation bubble cloud dynamics and surface erosion in high pressure fuel systems for medium/heavy duty Diesel engines)
1.1 Publishable summary
The primary objective of this MC fellowship is the characterisation of the effect bubble collapse processes on the Diesel nozzle flow and the induced erosion. To materialise this, a number of methodologies have been employed, ranging from fundamental single bubble dynamics models, RANS models simulating the nozzle flow under non-isothermal conditions including the heat generation/exchange with the nozzle walls. Finally, erosion predictions utilising various methodologies have been tested against available experimental data.
Initially, a critical assessment of experimental data has been performed that could assist the validation of the models resolving the bubble collapse dynamics near wall surfaces; the experimental data from NT University, Singapore, have been selected for validation of the relevant models. The predictive capability of various modelling and physical assumptions has been thoroughly assessed. More specifically, the Navier-Stokes equations expressed in 2D axisymmetric computational domain have been employed together with the VOF method for capturing the liquid-gas interface. Various assumptions with regards to the heat transfer between the gas inside the bubble and the surrounding liquid have been examined. Starting from isothermal conditions, polytropic and isentropic cases have been examined. Moreover, the energy equation has been also utilised to simulate the temperature during the collapse process, under the assumption of ideal gas behaviour. Finally, vaporisation/condensation of the liquid has been also included. The various finding support prior knowledge in this field; they have highlighted areas where experimental measurements are not easy or not accurate to be performed and have revealed the relative influence of the most influential parameters.
Turning to simulation of cavitation inside injector nozzles, a variety of methodologies have been also employed. Numerical simulations of the Navier-Stokes equations in 3D, including the motion of the needle valve have been performed, utilising in-house CFD codes. Initially, isothermal conditions and fixed values for the liquid properties have been assumed but challenged against the more complex cases of variable fuel properties (as function of pressure and temperature). For the latter, the heat generation due to wall friction inside the nozzle holes has been considered; the results have indicated a significant increase of temperature, which was also temporally and spatially resolved during the opening/closing of the injector’s needle valve. As an additional step, the assumption of adiabatic nozzle wall was challenged and the conjugate problem of fluid flow and wall heat transfer has been simulated. For that, appropriate boundary conditions resulting from nominal Diesel engine operating points have been utilised. Overall, many results have been obtained examining the effects of injection pressure, fuel properties, operating conditions and needle valve motion.
As a final step, predictions indicating the locations of erosion during the flow development inside Diesel injectors have been performed. These are based on various wall-erosion-indication functions, that consider the pressures induced during the collapse of cavitating bubbles. Different assumptions with regards to the numerical evaluation of these indicator functions have been obtained and guideline for the numerical implementation of best-performing ones have been derived. The results have been verified against experimental data for two injectors provided by Caterpillar Fuel Systems, US. Moreover, the effect of eccentric needle valve motion on erosion has been also evaluated for a VCO injector, showing significant differences.
Finally, during this period, the fellow had the opportunity to disseminate the work undertaken, participating in 6 conferences, while also preparing 2 papers for submission in peer-reviewed scientific journals.
The primary objective of this MC fellowship is the characterisation of the effect bubble collapse processes on the Diesel nozzle flow and the induced erosion. To materialise this, a number of methodologies have been employed, ranging from fundamental single bubble dynamics models, RANS models simulating the nozzle flow under non-isothermal conditions including the heat generation/exchange with the nozzle walls. Finally, erosion predictions utilising various methodologies have been tested against available experimental data.
Initially, a critical assessment of experimental data has been performed that could assist the validation of the models resolving the bubble collapse dynamics near wall surfaces; the experimental data from NT University, Singapore, have been selected for validation of the relevant models. The predictive capability of various modelling and physical assumptions has been thoroughly assessed. More specifically, the Navier-Stokes equations expressed in 2D axisymmetric computational domain have been employed together with the VOF method for capturing the liquid-gas interface. Various assumptions with regards to the heat transfer between the gas inside the bubble and the surrounding liquid have been examined. Starting from isothermal conditions, polytropic and isentropic cases have been examined. Moreover, the energy equation has been also utilised to simulate the temperature during the collapse process, under the assumption of ideal gas behaviour. Finally, vaporisation/condensation of the liquid has been also included. The various finding support prior knowledge in this field; they have highlighted areas where experimental measurements are not easy or not accurate to be performed and have revealed the relative influence of the most influential parameters.
Turning to simulation of cavitation inside injector nozzles, a variety of methodologies have been also employed. Numerical simulations of the Navier-Stokes equations in 3D, including the motion of the needle valve have been performed, utilising in-house CFD codes. Initially, isothermal conditions and fixed values for the liquid properties have been assumed but challenged against the more complex cases of variable fuel properties (as function of pressure and temperature). For the latter, the heat generation due to wall friction inside the nozzle holes has been considered; the results have indicated a significant increase of temperature, which was also temporally and spatially resolved during the opening/closing of the injector’s needle valve. As an additional step, the assumption of adiabatic nozzle wall was challenged and the conjugate problem of fluid flow and wall heat transfer has been simulated. For that, appropriate boundary conditions resulting from nominal Diesel engine operating points have been utilised. Overall, many results have been obtained examining the effects of injection pressure, fuel properties, operating conditions and needle valve motion.
As a final step, predictions indicating the locations of erosion during the flow development inside Diesel injectors have been performed. These are based on various wall-erosion-indication functions, that consider the pressures induced during the collapse of cavitating bubbles. Different assumptions with regards to the numerical evaluation of these indicator functions have been obtained and guideline for the numerical implementation of best-performing ones have been derived. The results have been verified against experimental data for two injectors provided by Caterpillar Fuel Systems, US. Moreover, the effect of eccentric needle valve motion on erosion has been also evaluated for a VCO injector, showing significant differences.
Finally, during this period, the fellow had the opportunity to disseminate the work undertaken, participating in 6 conferences, while also preparing 2 papers for submission in peer-reviewed scientific journals.