1、Advances in eddy-resolving computational approaches for flow and noise modellingSergey Karabasov,Queen Mary University of LondonAcknowledgements:Annabel Markesteijn,Hussain Abid,Igor Solntsev,Vasily GrayzevOutline of the talk Eddy-resolving simulations for flow and noise emissionsmodelling of rotati
2、ng machinery using state-of-the art commercial tools:advances and challenges for turbulence interaction problems,2 challenging use cases What is dissipation and dispersion errors and why one should care about them in scale resolving flow modelling?CABARET,out favourite low-dissipative and low-disper
3、sive method:some features and flow solution process Example 1:installed jet flow and noise modelling Example 2:installed propeller noise,from validation to design optimisation2Motivated by the search for transformative turbofanengine technologies which leverage a more compactengine-airframe integrat
4、ion to reduce drag(hence,fuelburnandemissions),whichposesnoiseproblems due to tightly coupled systemsSingle rotor-stator stage,R4 rotor with 22 blades and Outlet Guide Vanes(OGV),54=baseline and 26=low count(standard and reduced noise OGV blade design with a blade sweep)3 regimes:approach(61.7%fan s
5、peed),cutback(87.5%fan speed),and take-off(100%speed)Rotor wake/OGV interaction generates significant broadband(turbulence interaction noise)challenging for RANS-based prediction methods A particular challenge:prediction of the transonic fanregimes(cutback and take-off),where there is a strong shock
6、 appears in the rotor tip region,and multiple pure tones in the noise spectra,also referred to as“buzz-saw”noise.Use case 1:NASA Source Diagnostic Test3NASA/TM2003-212330Use case 1:the Wall Modelled LES accelerated on GPUs 4CharLES,Bres et al.AIAA 2024-3162Conventional 2nd order finite-volume scheme