東華大學圖書館 |

Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions./
作者:	Jordan, Cyrus Joshua.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2023,
面頁冊數:	146 p.
附註:	Source: Dissertations Abstracts International, Volume: 84-12, Section: A.
Contained By:	Dissertations Abstracts International84-12A.
標題:	Kinematics. -
電子資源:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30463916
ISBN:	9798379649579

Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions.
Jordan, Cyrus Joshua.

Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions. - Ann Arbor : ProQuest Dissertations & Theses, 2023 - 146 p.

Source: Dissertations Abstracts International, Volume: 84-12, Section: A.

Thesis (Ph.D.)--North Carolina State University, 2023.

This item must not be sold to any third party vendors.

The 21st century has sparked interest in hypersonic applications not seen since X-series planes were designed and tested in the early 1960s. These applications demand accurate and cost-effective tools for predicting and modeling hypersonic (M >5) conditions. Conventional wind tunnels cannot recreate hypersonic flight conditions limiting testing to specialized facilities. Shock and expansion tunnels provide high enthalpy conditions but are restricted to finite test time, while hypersonic blow-down tunnels exist but are costly and difficult to maintain. Often engineers turn to computational fluid dynamics (CFD) to simulate rather than test hypersonic applications. Turbulent flow is described by a broad spectrum of time and length scales, which must be resolved or modeled. Scale-resolving methods such as Direct Numerical Simulations (DNS) or Large Eddy Simulation (LES) require full or partial resolution of the turbulent scales. As hypersonic flows are described by Reynolds numbers (Re) of the order 106 per meter, and grid size requirements for DNS and LES roughly scale as Re3 and Re2 , scale-resolving methods often exceed computational resources. However, lower-fidelity methods model rather than resolve turbulence, with the current state-of-the-practice being Reynolds-averaged Navier Stokes (RANS) modeling.RANS predictions using Menter's baseline (BSL) and shear stress transport (SST) models show significant deviations from hypersonic experiments with shock wave boundary layer interactions (SBLIs), with BSL failing to predict axial separation and SST generally over-predicting the degree of separation. The LES-type methods improve upon these results, predicting the level of upstream influence more correctly for the stronger (more compressed) interactions while accurately capturing off-body structures in upstream boundary layers.In this work, we utilize high-fidelity datasets generated using scale-resolving methods (wallresolved LES or hybrid LES-RANS) to evaluate the model assumptions underpinning modern RANS practices. We show the Boussinesq hypothesis for the Reynolds shear stress and the gradient-diffusion assumption for the turbulent heat flux does not hold very well in the region of strong shock / boundary layer interaction but improve as the boundary layer reaches more of an equilibrium state. The extraction of turbulent momentum and thermal diffusivities reveals that the turbulent Prandtl number is not constant, reaching sub-unity values through much of the interaction region. The shear-stress structure factor (a key model constant in Menter's SST model) also varies significantly within the interaction region, rising above its equilibrium level near shock waves.Following this, a novel equation-learning neural net strategy was used to develop functional descriptions of turbulent Prandtl number and structure factor. Utilizing the high-fidelity hypersonic SBLI test suite and two additional, physically unique hypersonic test cases, we benchmark our developed models' performance. In general, for applications where SST over-predicts and BSL under-predicts experimental values, our modifications provided significant gains in accuracy in terms of separation and surface measurements. However, our model trended towards BSL predictions in weakly separated applications, typically over-predicting surface heat transfer, performing worse than unmodified SST. Overall, we have revealed physical inaccuracy in the fundamental assumptions underpinning RANS models in hypersonic SBLI applications. Our work demonstrates the potential of data-driven methods to develop model improvements with LES / LES-RANS simulations as truth models.

ISBN: 9798379649579Subjects--Topical Terms:

571109
Kinematics.

Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions.
LDR:04970nmm a2200409 4500 001 2397193
005 20240617111343.5
006 m o d
007 cr#unu||||||||
008 251215s2023 ||||||||||||||||| ||eng d
020 $a 9798379649579
035 $a (MiAaPQ)AAI30463916
035 $a (MiAaPQ)NCState_Univ18402040707
035 $a AAI30463916
040 $a MiAaPQ $c MiAaPQ
100 1 $a Jordan, Cyrus Joshua. $3 3766956
245 1 0 $a Turbulence Model Development for Hypersonic Shock Wave Boundary Layer Interactions.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2023
300 $a 146 p.
500 $a Source: Dissertations Abstracts International, Volume: 84-12, Section: A.
500 $a Advisor: Gremaud, Pierre;Narayanaswamy, Venkateswaran;Subbareddy, Pramod;Edwards, Jack.
502 $a Thesis (Ph.D.)--North Carolina State University, 2023.
506 $a This item must not be sold to any third party vendors.
520 $a The 21st century has sparked interest in hypersonic applications not seen since X-series planes were designed and tested in the early 1960s. These applications demand accurate and cost-effective tools for predicting and modeling hypersonic (M >5) conditions. Conventional wind tunnels cannot recreate hypersonic flight conditions limiting testing to specialized facilities. Shock and expansion tunnels provide high enthalpy conditions but are restricted to finite test time, while hypersonic blow-down tunnels exist but are costly and difficult to maintain. Often engineers turn to computational fluid dynamics (CFD) to simulate rather than test hypersonic applications. Turbulent flow is described by a broad spectrum of time and length scales, which must be resolved or modeled. Scale-resolving methods such as Direct Numerical Simulations (DNS) or Large Eddy Simulation (LES) require full or partial resolution of the turbulent scales. As hypersonic flows are described by Reynolds numbers (Re) of the order 106 per meter, and grid size requirements for DNS and LES roughly scale as Re3 and Re2 , scale-resolving methods often exceed computational resources. However, lower-fidelity methods model rather than resolve turbulence, with the current state-of-the-practice being Reynolds-averaged Navier Stokes (RANS) modeling.RANS predictions using Menter's baseline (BSL) and shear stress transport (SST) models show significant deviations from hypersonic experiments with shock wave boundary layer interactions (SBLIs), with BSL failing to predict axial separation and SST generally over-predicting the degree of separation. The LES-type methods improve upon these results, predicting the level of upstream influence more correctly for the stronger (more compressed) interactions while accurately capturing off-body structures in upstream boundary layers.In this work, we utilize high-fidelity datasets generated using scale-resolving methods (wallresolved LES or hybrid LES-RANS) to evaluate the model assumptions underpinning modern RANS practices. We show the Boussinesq hypothesis for the Reynolds shear stress and the gradient-diffusion assumption for the turbulent heat flux does not hold very well in the region of strong shock / boundary layer interaction but improve as the boundary layer reaches more of an equilibrium state. The extraction of turbulent momentum and thermal diffusivities reveals that the turbulent Prandtl number is not constant, reaching sub-unity values through much of the interaction region. The shear-stress structure factor (a key model constant in Menter's SST model) also varies significantly within the interaction region, rising above its equilibrium level near shock waves.Following this, a novel equation-learning neural net strategy was used to develop functional descriptions of turbulent Prandtl number and structure factor. Utilizing the high-fidelity hypersonic SBLI test suite and two additional, physically unique hypersonic test cases, we benchmark our developed models' performance. In general, for applications where SST over-predicts and BSL under-predicts experimental values, our modifications provided significant gains in accuracy in terms of separation and surface measurements. However, our model trended towards BSL predictions in weakly separated applications, typically over-predicting surface heat transfer, performing worse than unmodified SST. Overall, we have revealed physical inaccuracy in the fundamental assumptions underpinning RANS models in hypersonic SBLI applications. Our work demonstrates the potential of data-driven methods to develop model improvements with LES / LES-RANS simulations as truth models.
590 $a School code: 0155.
650 4 $a Kinematics. $3 571109
650 4 $a Finite volume method. $3 1027193
650 4 $a Fluid dynamics. $3 545210
650 4 $a Jet engines. $3 3698604
650 4 $a Pressure distribution. $3 3564852
650 4 $a Energy. $3 876794
650 4 $a Boundary conditions. $3 3560411
650 4 $a Friction. $3 650299
650 4 $a Heat transfer. $3 3391367
650 4 $a Physics. $3 516296
650 4 $a Partial differential equations. $3 2180177
650 4 $a Viscosity. $3 1050706
650 4 $a Gases. $3 559387
650 4 $a Turbulence models. $3 3682363
650 4 $a Reynolds number. $3 3681436
650 4 $a Geometry. $3 517251
650 4 $a Shear stress. $3 3681838
650 4 $a Aerospace engineering. $3 1002622
650 4 $a Fluid mechanics. $3 528155
650 4 $a Mathematics. $3 515831
650 4 $a Mechanics. $3 525881
650 4 $a Thermodynamics. $3 517304
690 $a 0791
690 $a 0605
690 $a 0538
690 $a 0204
690 $a 0338
690 $a 0405
690 $a 0346
690 $a 0348
710 2 $a North Carolina State University. $3 1018772
773 0 $t Dissertations Abstracts International $g 84-12A.
790 $a 0155
791 $a Ph.D.
792 $a 2023
793 $a English
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30463916