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

PIV measurements in a turbulent boundary layer
overlying a spanwise heterogeneous roughness

R. Yao1, K. T. Christensen1,2

1 University of Notre Dame, Department of Aerospace and Mechanical Engineering, Notre Dame, IN,
46556, USA

2 Illinois Institute of Technology, Department of Mechanical, Materials and Aerospace Engineering &
Civil, Architectural and Environmental Engineering, Chicago, IL, 60616, USA

1 Introduction
In nature and engineering applications, wall-bounded flow often encounter a heterogeneous surface condi-
tion, such as the atmosphere boundary layer at the urban boundary and flow over riveted aircraft surfaces. In
a particular scenario, when the surface heterogeneity is predominantly in the spanwise direction of the flow,
this roughness heterogeneity can generate secondary flow in cross flow plane which is very different from
smooth-wall or homogeneous rough-wall boundary layers.

Figure 1: (a) Important parameters of the ridge-type roughness: h, ridge height; S, the spacing of the ridges;
W , ridge width. (b) Surface models used in these experiments (the RI of the clear one is matched with that
of the working fluid during experiments). (c) Planar PIV setup for streamwise–wall-normal measurements.

A number of experimental (Barros and Christensen, 2014; Vanderwel and Ganapathisubramani, 2015;
Zampiron et al., 2020) and numerical (Anderson et al., 2015; Hwang and Lee, 2018; Castro et al., 2021) stud-
ies have shown features associated with the secondary flow including: streamwise oriented alternating low-
and high-momentum pathways (LMP, HMP); downwash and upwash flow at HMP and LMP, respectively;
counter-rotating streamwise vortices formed across an LMP and HMP. The secondary flow mechanism is
explained based on Prandtl’s secondary flow of the second kind according to Anderson et al. (2015). With
this current understanding, this work uses PIV to document the spatio-temporal information of this special
flow in x− y and y− z planes over a ridged spanwise heterogeneous rough surface. The results present here
mainly focus on the mean and turbulence statistics measured in a flow facility at moderate Reynolds number.

2 Methods
The roughness models used in the current experiments are derived from an idealized ridge-type roughness,
wherein the roughness topography is constructed from equally spaced bars of square cross-section (Fig 1a).
In the design of the roughness model, the size and spacing of the ridge elements were selected to ensure
the generation of strong secondary flows for ridged roughness (Vanderwel and Ganapathisubramani, 2015;
Medjnoun et al., 2018). PIV measurements are performed in a refractive-index-matched (RIM) flow facility
(Fig. 1c) wherein the refractive index of the working fluid (aqueous NaI) matches that of solid models of the
roughness so that reflections and aberrations in the PIV illumination and imaging are minimized and flow in
the immediate vicinity of the roughness can be interrogated by PIV. Figure 1(b) shows the roughness tiles
for RIM and the overall roughness arrangement in the test section.



Figure 2: (a) Ensemble-averaged streamwise velocity field between at the center between spanwise-
separated ridges. (b,c) Ensemble and streamwise-averaged streamwise and wall-normal velocity profiles
positioned on top of a ridge and between two ridges.

3 Results
Initial results showing the time-averaged and time and streamwise-averaged velocities from x−y plane PIV
measurements are shown in Fig. 2. Figure 2a shows the time-averaged streamwise component velocity in
the position between two ridges from wall to the free stream. In this case the data below y = 3.8 mm is
behind the ridges and therefore demonstrates the applicability of RIM in this case. Figure 2a also verifies
the streamwise homogeneity of the mean flow. The ensemble- and streamwise-averaged velocity profiles
are calculated and shown in Figs. 2b,c. The profiles confirm that the HMP and downwash region is located
in between the ridges while the LMP and upwash is observed above the ridges which is consistent with
observations in the literature (Vanderwel and Ganapathisubramani, 2015; Medjnoun et al., 2018; Castro
et al., 2021). The ongoing progress of this work includes: stereo-PIV measurements in the cross-plane,
frequency analysis, and quantification of inner–outer interactions in this flow scenario. Follow-on research
will focus on aligning these same ridges obliquely to the mean flow to understand the persistence of these
secondary flows relative to the spanwise heterogeneity of the topography.

References
Anderson W, Barros JM, Christensen KT, and Awasthi A (2015) Numerical and experimental study of mech-

anisms responsible for turbulent secondary flows in boundary layer flows over spanwise heterogeneous
roughness. Journal of Fluid Mechanics 768:316–347

Barros JM and Christensen KT (2014) Observations of turbulent secondary flows in a rough-wall boundary
layer. Journal of Fluid Mechanics 748

Castro IP, Kim J, Stroh A, and Lim HC (2021) Channel flow with large longitudinal ribs. Journal of Fluid
Mechanics 915

Hwang HG and Lee JH (2018) Secondary flows in turbulent boundary layers over longitudinal surface
roughness. Physical Review Fluids 3:014608

Medjnoun T, Vanderwel C, and Ganapathisubramani B (2018) Characteristics of turbulent boundary layers
over smooth surfaces with spanwise heterogeneities. Journal of Fluid Mechanics 838:516–543

Vanderwel C and Ganapathisubramani B (2015) Effects of spanwise spacing on large-scale secondary flows
in rough-wall turbulent boundary layers. Journal of Fluid Mechanics 774

Zampiron A, Cameron S, and Nikora V (2020) Secondary currents and very-large-scale motions in open-
channel flow over streamwise ridges. Journal of Fluid Mechanics 887


	Introduction
	Methods
	Results