| Autor: |
Antonio Agresta, Alessandro Anderlini, Luca Matteucci, Maria Vittoria Salvetti |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
Lecture Notes in Computational Science and Engineering ISBN: 9783030487201 |
| DOI: |
10.1007/978-3-030-48721-8_9 |
| Popis: |
A stochastic sensitivity analysis and calibration of the cavitation model parameters in the URANS simulations of a configuration representative of high-pressure injectors for automotive applications is carried out. A popular homogeneous-flow cavitation model is considered, in which the mass transfer due to cavitation is given by the Schnerr–Sauer model together with the classical Rayleigh–Plesset equation. A stochastic approach based on the generalized Polynomial Chaos (gPC) expansion is adopted, which allows continuous response surfaces of the quantities of interest in the parameter space to be obtained starting from a few deterministic simulations. The considered uncertain parameters are the so-called scaling factors. The calibration of these parameters is carried out by using the gPC response surfaces for a axisymmetric simplified geometry against the experimental value of the critical cavitation point, i.e. the condition at which the injector is choked. The procedure is carried out for two different turbulence models, viz. the k − ω SST and RSM models. The so-obtained optimal parameter set-ups are then validated for the real three-dimensional geometry. The k − ω SST optimal set-up gives very accurate predictions also in the three-dimensional case. Finally, the results obtained with this optimal set-up are compared to those given by standard values, confirming that the predictions of the different flow regimes occurring in high-pressure injectors are highly sensitive to cavitation model parameters. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
