Proper orthogonal decomposition analysis of a circular cylinder undergoing vortex-induced vibrations

Riches, Graham; Martinuzzi, Robert J.; Morton, Chris

doi:10.1063/1.5046090

Cited by 82 publications

(21 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2005; Perrin et al. 2007; Riches, Martinuzzi & Morton 2018). However, as seen in figure 15( f ), the radius of the orbits, defined as , varies in time.…”

Section: Resultsmentioning

confidence: 99%

Flows over surface-mounted bluff bodies with different spanwise widths submerged in a deep turbulent boundary layer

Fang

Tachie

2019

J. Fluid Mech.

View full text Add to dashboard Cite

The spatio-temporal dynamics of separation bubbles induced by surface-mounted bluff bodies with different spanwise widths and submerged in a thick turbulent boundary layer is experimentally investigated. The streamwise extent of the bluff bodies is fixed at 2.36 body heights and the spanwise aspect ratio ($AR$), defined as the ratio between the width and height, is increased from 1 to 20. The thickness of the upstream turbulent boundary layer is 4.8 body heights, and the dimensionless shear and turbulence intensity evaluated at the body height are 0.23 % and 15.8 %, respectively, while the Reynolds number based on the body height and upstream free-stream velocity is 12 300. For these upstream conditions and limited streamwise extent of the bluff bodies, two distinct and strongly interacting separation bubbles are formed over and behind the bluff bodies. A time-resolved particle image velocimetry is used to simultaneously measure the velocity field within these separation bubbles. Based on the dynamics of the mean separation bubbles over and behind the bluff bodies, the flow fields are categorized into three-dimensional, transitional and two-dimensional regimes. The results indicate that the low-frequency flapping motions of the separation bubble on top of the bluff body with $\mathit{AR}=1$ are primarily influenced by the vortex shedding motion, while those with larger aspect ratios are modulated by the large-scale streamwise elongated structures embedded in the oncoming turbulent boundary layer. For $\mathit{AR}=1$ and 20, the flapping motions in the wake region are strongly influenced by those on top of the bluff bodies but with a time delay that is dependent on the $AR$. Moreover, an expansion of the separation bubble on the top surface tends to lead to an expansion and contraction of separation bubbles in the wake of $\mathit{AR}=20$ and 1, respectively. As for the transitional case of $\mathit{AR}=8$, the separation bubbles over and behind the body are in phase over a wide range of time difference. The dynamics of the shear layer in the wake region of the transitional case is remarkably more complex than the limiting two-dimensional and three-dimensional configurations.

show abstract

“…2005; Perrin et al. 2007; Riches, Martinuzzi & Morton 2018). However, as seen in figure 15( f ), the radius of the orbits, defined as , varies in time.…”

Section: Resultsmentioning

confidence: 99%

Flows over surface-mounted bluff bodies with different spanwise widths submerged in a deep turbulent boundary layer

Fang

Tachie

2019

J. Fluid Mech.

View full text Add to dashboard Cite

show abstract

“…How should we measure and interpret airfoil stall behavior? It should be clear given the discussion so far that simple averaging or even phase averaging will remove important data (Riches et al, 2018). The cycle to cycle variations and outliers are an important part of the dataset and shouldn't be smeared out.…”

Section: So How Are These Variations Treated?mentioning

confidence: 99%

“…In particular, there is abundant literature using Singular Value Decomposition (SVD) based methods such as POD/PCA (Taira et al, 2017), DMD (Schmid, 2010;Kutz et al, 2015;, SPOD (Sieber et al, 2015). These methods generally do not perform well in cases with any kind of traveling wave behavior (Taira et al, 2017;Riches et al, 2018;Hosseini et al, 2016). The reason for this lies in the creation of fixed spatial functions/basis functions.…”

Section: So How Are These Variations Treated?mentioning

confidence: 99%

Cartographing dynamic stall with machine learning

Lennie

Steenbuck

Noack

et al. 2019

Preprint

View full text Add to dashboard Cite

Abstract. Airfoil stall is bad for wind turbines. Once stall has set in, lift collapses, drag increases and then both of these forces will fluctuate strongly. The result is higher fatigue loads and lower energy yield. In dynamic stall, separation first develops from the trailing edge up the leading edge, eventually the shear layer rolls up and then a coherent vortex forms and then sheds downstream with it’s low pressure core causing a lift spike and moment dump. When 50+ experimental cycles of lift or pressure values are averaged, this process appears clear and coherent in flow visualizations. Unfortunately, stall is not one clean process, but a broad collection of processes. This means that the analysis of separated flows should be able to detect outliers and analysis cycle to cycle variations. Modern data science/machine learning can be used to treat separated flows. In this study, a clustering method based on dynamic time warping is used to find different shedding behaviors. This method captures that secondary and tertiary vorticity vary strongly and in static stall with surging flow; the flow can occasionally reattach. A convolutional neural network was used to extract dynamic stall vorticity convection speeds and phases from pressure data. Finally, bootstrapping was used to provide best practices regarding the number of experimental repetitions required to ensure experimental convergence.

show abstract

“…This means that modeling separated flows has always been a challenging part of designing wind turbines or even understanding experimental and field data. Even in the age of computational fluid dynamics (CFD), attempts to simulate stall with unsteady Reynolds-averaged Navier-Stokes (URANS) equations have not yet yielded good-quality results (Strangfeld et al, 2015;Rumsey, 2008;Rumsey and Nishino, 2011). Large-eddy simulations (LESs) show promise but are still too computationally expensive to be used as an ordinary design and analysis tool (Rumsey and Nishino, 2011).…”

Section: Introductionmentioning

confidence: 99%

Cartographing dynamic stall with machine learning

et al. 2020

View full text Add to dashboard Cite

Abstract. Once stall has set in, lift collapses, drag increases and then both of these forces will fluctuate strongly. The result is higher fatigue loads and lower energy yield. In dynamic stall, separation first develops from the trailing edge up the leading edge. Eventually the shear layer rolls up, and then a coherent vortex forms and then sheds downstream with its low-pressure core causing a lift overshoot and moment drop. When 50+ experimental cycles of lift or pressure values are averaged, this process appears clear and coherent in flow visualizations. Unfortunately, stall is not one clean process but a broad collection of processes. This means that the analysis of separated flows should be able to detect outliers and analyze cycle-to-cycle variations. Modern data science and machine learning can be used to treat separated flows. In this study, a clustering method based on dynamic time warping is used to find different shedding behaviors. This method captures the fact that secondary and tertiary vorticity vary strongly, and in static stall with surging flow the flow can occasionally reattach. A convolutional neural network was used to extract dynamic stall vorticity convection speeds and phases from pressure data. Finally, bootstrapping was used to provide best practices regarding the number of experimental repetitions required to ensure experimental convergence.

show abstract

Proper orthogonal decomposition analysis of a circular cylinder undergoing vortex-induced vibrations

Cited by 82 publications

References 42 publications

Flows over surface-mounted bluff bodies with different spanwise widths submerged in a deep turbulent boundary layer

Flows over surface-mounted bluff bodies with different spanwise widths submerged in a deep turbulent boundary layer

Cartographing dynamic stall with machine learning

Cartographing dynamic stall with machine learning

Contact Info

Product

Resources

About