kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
In situ visualization of large-scale turbulence simulations in Nek5000 with ParaView Catalyst
KTH, Centres, SeRC - Swedish e-Science Research Centre. KTH, School of Engineering Sciences (SCI), Centres, Linné Flow Center, FLOW.ORCID iD: 0000-0003-0790-8460
KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Computational Science and Technology (CST).ORCID iD: 0000-0002-3234-9368
KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Theoretical Computer Science, TCS. KTH, School of Electrical Engineering and Computer Science (EECS), Computer Science, Computational Science and Technology (CST).ORCID iD: 0000-0001-6408-3333
KTH, School of Engineering Sciences (SCI), Engineering Mechanics.ORCID iD: 0000-0002-6712-8944
Show others and affiliations
2022 (English)In: Journal of Supercomputing, ISSN 0920-8542, E-ISSN 1573-0484, Vol. 78, no 3, p. 3605-3620Article in journal (Refereed) Published
Abstract [en]

In situ visualization on high-performance computing systems allows us to analyze simulation results that would otherwise be impossible, given the size of the simulation data sets and offline post-processing execution time. We develop an in situ adaptor for Paraview Catalyst and Nek5000, a massively parallel Fortran and C code for computational fluid dynamics. We perform a strong scalability test up to 2048 cores on KTH’s Beskow Cray XC40 supercomputer and assess in situ visualization’s impact on the Nek5000 performance. In our study case, a high-fidelity simulation of turbulent flow, we observe that in situ operations significantly limit the strong scalability of the code, reducing the relative parallel efficiency to only ≈ 21 % on 2048 cores (the relative efficiency of Nek5000 without in situ operations is ≈ 99 %). Through profiling with Arm MAP, we identified a bottleneck in the image composition step (that uses the Radix-kr algorithm) where a majority of the time is spent on MPI communication. We also identified an imbalance of in situ processing time between rank 0 and all other ranks. In our case, better scaling and load-balancing in the parallel image composition would considerably improve the performance of Nek5000 with in situ capabilities. In general, the result of this study highlights the technical challenges posed by the integration of high-performance simulation codes and data-analysis libraries and their practical use in complex cases, even when efficient algorithms already exist for a certain application scenario.

Place, publisher, year, edition, pages
Springer , 2022. Vol. 78, no 3, p. 3605-3620
Keywords [en]
Computational fluid dynamics, High-performance computing, In situ visualization, Catalysts, Data visualization, Efficiency, Image enhancement, Scalability, Supercomputers, Visualization, Application scenario, High performance computing systems, High-fidelity simulations, High-performance simulation, Large scale turbulence, Parallel efficiency, Relative efficiency, Technical challenges, In situ processing
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:kth:diva-311178DOI: 10.1007/s11227-021-03990-3ISI: 000680293400003PubMedID: 35210696Scopus ID: 2-s2.0-85111797526OAI: oai:DiVA.org:kth-311178DiVA, id: diva2:1655482
Note

QC 20220502

Available from: 2022-05-02 Created: 2022-05-02 Last updated: 2024-01-19Bibliographically approved
In thesis
1.
The record could not be found. The reason may be that the record is no longer available or you may have typed in a wrong id in the address field.
2. Wings, turbulent boundary layers and flow separation
Open this publication in new window or tab >>Wings, turbulent boundary layers and flow separation
2024 (English)Doctoral thesis, comprehensive summary (Other academic)
Alternative title[sv]
Vingar, turbulenta gränsskikt och avlösning
Abstract [en]

The present doctoral thesis investigates the turbulent flow developing around wing sections, focusing on the impact of adverse-pressure-gradient (APG) conditions on turbulent boundary layers (TBLs) and the physics of flow separation. Both experimental and numerical methods are employed to generate high-fidelity data sets and provide an in-depth analysis of the flow.

The first objective of this thesis is the development of a comprehensive database for the flow around a NACA 4412 wing profile. For this purpose, adaptive mesh refinement (AMR) is used together with the spectral-element method code Nek5000. With AMR, high-resolution Large Eddy Simulations (LES) are conducted at various Reynolds numbers (Rec = 2×105, 4×105 and 1×106) and angles of attack (AoA=5°, 8°, 11°, 14°), which were previously unattainable. The effect that strong APGs have on TBLs developing around a wing section is assessed through the collection of statistics and time series. The results demonstrate the influence of APG conditions on both the mean and variance profiles of velocity, and on the distribution and production of turbulence energy within the TBL. Additionally, the connection of APG TBLs with flow separation is explored through the development of an in-situ identification and tracking algorithm, tightly integrated into Nek5000. Our findings show that, in contrast to canonical flows, backflow events in TBLs under strong APGs extensively merge to form larger structures that grow exponentially in size, eventually leading to significant flow separation near the wing’s trailing edge.

Furthermore, a wind-tunnel experimental campaign is conducted to validate and extend the numerical results. Pressure, wall-shear stress and velocity measurements were carried out in the MTL wind tunnel at KTH Royal Institute of Technology. The study also scrutinizes measurement methodologies for APG TBLs, examining uncertainties in skin-friction determination and the impact of hot-wire probe lengths on velocity variance profiles.

Finally, a study based on Reynolds-averaged Navier–Stokes (RANS) simulations, utilizing high-fidelity data for validation, is performed to assess the optimization of flow-control schemes based on blowing and suction. This study, later extended to a transonic airfoil, showcases Bayesian optimization (BO) as an efficient method for computational fluid dynamics (CFD)-based optimization problems.

Abstract [sv]

Denna doktorsavhandling undersöker det turbulenta flödet runt vingsektioner, med fokus på effekten av negativa tryckgradienter (APG) på turbulenta gränsskikt (TBL) och fysiken bakom avlösning. Både experimentella och numeriska metoder används för att generera data och genomföra en noggrant analys av flödet.

Det första målet med avhandlingen är att utveckla en omfattande databas för flödet runt en NACA 4412 vingprofil. För detta ändamål används adaptiv nätförfining (AMR) i det spektralelementbaserade programmet Nek5000. Med AMR genomförs väggupplösta large-eddy simuleringar (LES) vid olika Reynoldstal (Rec = 2×105, 4×105 och 1×106) och anfallsvinklar (AoA=5°, 8°, 11°, 14°), vilka tidigare varit ouppnåeliga. Effekten av den starka negativa tryckgradienten på TBL som utvecklas runt vingsektionen bedöms genom insamling av statistik och tidsserier. Resultaten visar på APG:s påverkan på både medel- och variansprofiler för hastighet, samt på fördelning och produktion av turbulenta energien inom TBL. Dessutom utforskas sambandet mellan APG TBL och avlösning genom utveckling av en in-situ identifierings- och spårningsalgoritm, integrerad i Nek5000. Våra resultat visar att negativa hastigheter händelser i TBL under starka APG interagerar betydligt med varandra, sammanflätar och bildar större strukturer som ökar exponentiellt i storlek, och så småningom leder till betydande avlösning nära vingens bakkant. 

Dessutom genomförs en experimentkampanj i vindtunnel för att validera och utöka de numeriska resultaten. Mätningar av tryck, väggskjuvspänning och hastighet utförs i MTL vindtunneln vid KTH Kungliga Tekniska Högskolan. Studien granskar också mätmetoder för APG TBL, undersöker osäkerheter i bestämningen av väggskjuvspänning samt effekterna av längden på varmtrådprober på mätprofiler för hastighetsvarians.

Slutligen utförs en RANS-studie, där högupplösta data används för validering, för att bedöma optimering av flödeskontrollmetoder baserade på blåsning och sugning. Denna studie, som senare utvidgas till ett transoniskt vingprofil, visar på Bayesiansk optimering som en effektiv metod för CFD (computational fluid dynamics)-baserade optimeringsproblem.

Place, publisher, year, edition, pages
Stockholm: KTH Royal Institute of Technology, 2024. p. 101
Series
TRITA-SCI-FOU ; 2024-03
Keywords
Wings, turbulent boundary layers, adverse pressure gradients, flow separation, numerical simulations, wind-tunnel experiments, flow control., Vingar, turbulenta gränsskikt, negativa tryckgradienter, avlösning, numeriska simuleringar, vindtunnel-experiment, flow control
National Category
Fluid Mechanics
Research subject
Aerospace Engineering; Engineering Mechanics
Identifiers
urn:nbn:se:kth:diva-342447 (URN)978-91-8040-827-1 (ISBN)
Public defence
2024-02-16, F3, Lindstedtsvägen 26, Stockholm, 10:15 (English)
Opponent
Supervisors
Funder
Knut and Alice Wallenberg Foundation
Note

QC 240122

Available from: 2024-01-22 Created: 2024-01-19 Last updated: 2025-02-09Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textPubMedScopus

Authority records

Atzori, MarcoKöpp, WiebkeChien, Wei DerMassaro, DanieleMallor, FerminPeplinski, AdamRezaei, MohammadtaghiJansson, NiclasMarkidis, StefanoVinuesa, RicardoSchlatter, PhilippWeinkauf, Tino

Search in DiVA

By author/editor
Atzori, MarcoKöpp, WiebkeChien, Wei DerMassaro, DanieleMallor, FerminPeplinski, AdamRezaei, MohammadtaghiJansson, NiclasMarkidis, StefanoVinuesa, RicardoSchlatter, PhilippWeinkauf, Tino
By organisation
SeRC - Swedish e-Science Research CentreLinné Flow Center, FLOWComputational Science and Technology (CST)Theoretical Computer Science, TCSEngineering MechanicsCentre for High Performance Computing, PDC
In the same journal
Journal of Supercomputing
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
pubmed
urn-nbn

Altmetric score

doi
pubmed
urn-nbn
Total: 647 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf