14th International Symposium on Particle Image Velocimetry – ISPIV 2021
August 1–5, 2021

Combining high-speed planar PIV and motion
tracking of a flexible cylinder in cross-flow

D. G. Gundersen1, K. T. Christensen1,2, G. Blois1

1 University of Notre Dame, Department of Aerospace and Mechanical Engineering, Notre Dame, USA
2 Illinois Institute of Technology, Depts. of Mechanical, Materials & Aerospace Engineering and Civil,

Architectural & Environmental Engineering, Chicago, USA

1 Introduction
Most modeling studies investigating the flow dynamics in vegetation canopies are limited to rigid models
as proxies for vegetation elements. However, most canopies embody some degree of structural flexibility,
resulting in aeroelastic mechanisms coupling the motion of the vegetation with the surrounding flow. Stud-
ies addressing flexible canopies typically quantify either the flow or the plant motion independently, thus
missing the instantaneous coupling between turbulent stresses and structural deformations. Few experiments
have been devoted to measuring both quantities simultaneously. Okamoto and Nezu (2009) utilized a com-
bined PIV-PTV technique to capture both flow and canopy motion. However, only the motion of the stem
tips was captured, as opposed to the deformation of the entire stem. Py et al. (2006) employed digital image
correlation (DIC) to quantify the motion of crop canopies using in-field images. However, the wind itself
was not measured across the domain.

The present work presents an experimental technique that can be utilized to study the flow–structure in-
teraction in flexible canopies, and that could be extended to other flexible and/or moving objects. High-speed
PIV data of the flow surrounding an idealized canopy element, consisting of a flexible cylinder, together with
the corresponding displacement field throughout the cylinder were simultaneously obtained combining flu-
orescent imaging and refractive index matching (RIM).

2 Methods
Laboratory experiments were performed in the Refractive Index-Matched (RIM) flow facility at Notre
Dame (Blois et al., 2020). The RIM approach involved using an aqueous solution of sodium iodide (NaI) as
the working fluid and solid models whose RI matches that of the fluid, thus obtaining an optical continuum
between the solid model and the surrounding fluid. The operating principle of the technique presented herein
involved seeding the two phases with different tracers, facilitating image segmentation and interrogation of
each effect independently. The working fluid was seeded with silver-coated particles with a 14-micron di-
ameter and a specific gravity of 1.7. The cylinder models were fabricated from a polyurethane rubber with
an index of refraction of n = 1.4865 and a Young’s Modulus of E = 2.9 MPa. A small volume of orange flu-
orescent dye was embedded within the polyurethane during the casting process at a concentration of 1µL/g.
The casting protocol was refined over multiple iterations to maximize particle dispersion uniformity and to
optimize dye-to-solid ratio. The cylinder was mounted on the bottom wall of the test section with its axis
oriented in the normal direction. The cylinder had a length of L = 43 mm and a diameter of D = 4.5 mm.
The model was fully immersed in the boundary layer, with a thickness of δ99 = 58.4 mm. The laser sheet
was oriented in the streamwise–wall-normal plane intersecting the center of the cylinder. All particles were
illuminated by the same laser sheet formed from a 50 mJ, dual-head Nd:YLF high-speed laser. A single
high-speed digital camera was available for these tests. The camera captured both the light scattered by the
silver-coated particles seeding the fluid flow and the light emitted by the particles seeded within the cylinder.
The signal was separated using robust segmentation algorithms. The displacement field at any point in the
flow and within the cylinder was obtained by standard cross-correlation schemes. The segmentation scheme
was automated, allowing for the very large datasets produced by the high-speed imaging arrangement to be
processed.



(a)

𝑈 (m/s)

(b)

Figure 1: (a) Instantaneous snapshot of the streamwise velocity, U , with vectors displaying the fluctuating
velocities. The cylinder location is shown in gray. (b) Instantaneous displacement vectors relative to the
previous image. For better visualization, the vector magnitudes are exaggerated by a factor of 10, and a
reference vector of 0.5 mm is shown in the top left.

3 Results
Time-resolved planar PIV data was acquired with a sample frequency of 640 image pairs per second and a
sample size of 2734 image pairs. The flow data were processed with an iterative multi-pass interrogation
method, resulting in a vector grid spacing of 0.33 mm. A representative example of the instantaneous distri-
bution of streamwise velocity, U , around the cylinder model subjected to an oncoming turbulent boundary
layer is shown in Fig. 1(a). The results shows that, despite a slight decrease in illumination energy in the
shadow region underneath the cylinder, the data quality is relatively uniform throughout the field of view.
Fig. 1(b) shows the cylinder displacement map at a given snapshot. The vectors are oriented perpendicular to
the model centerline and the vector magnitudes increase with distance from the base. These observations are
qualitatively consistent with the expected displacement distribution of a cantilever beam that is fixed to the
wall and subjected to an oncoming turbulent boundary layer. This experimental technique allowed for the
accurate calculation of quantities such as power spectra of the cylinder vibrations and the vortex shedding
frequency. In addition, it also facilitated temporal correlations between the cylinder and flow dynamics.

4 Conclusions
The preliminary results presented herein demonstrate a technique to successfully embed fluorescent particles
in a transparent flexible vegetation model. The utilization of this model in a RIM environment allows one to
quantify the motion and deformation of the illuminated object subjected to a turbulent flow. High-speed PIV
data allowed for the observation of the dynamic link between a deformable object and the surrounding flow,
thus giving the ability to probe the fluid–structure coupling. The experimental method could be utilized to
aid in the study of other aeroelastic flow–structure interactions.

References
Blois G, Bristow NR, Kim T, Best JL, and Christensen KT (2020) Novel environment enables piv mea-

surements of turbulent flow around and within complex topographies. Journal of Hydraulic Engineering
146:04020033

Okamoto TA and Nezu I (2009) Turbulence structure and “monami” phenomena in flexible vegetated open-
channel flows. J Hydraul Res 47:798–810

Py C, De Langre E, and Moulia B (2006) A frequency lock-in mechanism in the interaction between wind
and crop canopies. J Fluid Mech 568:425–449


	Introduction
	Methods
	Results
	Conclusions