Scholarly article on topic 'Similarities and Differences between Hot-film Anemometry and Particle Image Velocimetry in Open Channels'

Similarities and Differences between Hot-film Anemometry and Particle Image Velocimetry in Open Channels Academic research paper on "Chemical engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Procedia Engineering
OECD Field of science
Keywords
{"Hot-Fim Anemometry" / "Patricle Image Velocimetry" / "Velocity Distribution" / "Turbulent Characteristics ;Open Channel ;"}

Abstract of research paper on Chemical engineering, author of scientific article — Evangelos Keramaris

Abstract In this study a comparison between hot-film anemometry and Particle Image Velocimetry (PIV) measurements in open channels is investigated. For this reason individual experiments with these two methods were conducted. Initially, experiments were performed with hot-film anemometry method and then with PIV method in open channels with the same conditions. The comparison is curried out for two cases of study: a) impermeable bed and b) permeable bed. For the simulation of the permeable bed a porous filter with porous thicknesss’=2cm and porosity ɛ=0.75 was used. Sixteen (16) laboratory experiments were carried out, eight (8) for each case. Velocity distribution and turbulent characteristics of the flow above the impermeable bed and above the porous filters for all the different total water heights and for the same channel slope were evaluated. Results show that the different method for experimental measurements influence in some cases the turbulent characteristics of the flow, especially in the case of permeable bed. Also the results show that there is a good agreement between these two methods, both for permeable and impermeable bed, regarding the velocity distribution.

Academic research paper on topic "Similarities and Differences between Hot-film Anemometry and Particle Image Velocimetry in Open Channels"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedía Engineering 162 (2016) 388 - 395

Procedía Engineering

www.elsevier.com/locate/procedia

International Conference on Efficient & Sustainable Water Systems Management toward Worth

Living Development, 2nd EWaS 2016

Similarities and Differences Between Hot-Film Anemometry and Particle Image Velocimetry in Open Channels

Evangelos Keramarisa'*

aUniversity of Thessaly, Department of Civil Engineering, Pedion Areos, 38334 Volos, Greece

Abstract

In this study a comparison between hot-film anemometry and Particle Image Velocimetry (PIV) measurements in open channels is investigated. For this reason individual experiments with these two methods were conducted. Initially, experiments were performed with hot-film anemometry method and then with PIV method in open channels with the same conditions.The comparison is curried out for two cases of study: a) impermeable bed and b) permeable bed. For the simulation of the permeable bed a porous filter with porous thicknesss'=2cm and porosity s=0.75 was used. Sixteen (16) laboratory experiments were carried out, eight (8) for each case. Velocity distribution and turbulent characteristics of the flow above the impermeable bed and above the porous filters for all the different total water heights and for the same channel slope were evaluated. Results show that the different method for experimental measurements influence in some cases the turbulent characteristics of the flow, especially in the case of permeable bed. Also the results show that there is a good agreement between these two methods, both for permeable and impermeable bed, regarding the velocity distribution.

©2016 The Authors.PublishedbyElsevierLtd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review underresponsibilityoftheorganizingcommitteeoftheEWaS2 International Conference on Efficient & Sustainable Water Systems Management toward Worth Living Development

Keywords. Hot-Fim Anemometry; Patríele Image Velocimetry; Velocity Distribution; Turbulent Characteristics;Open Channel;

* Corresponding author. Tel.: +302421074140. E-mail address: ekeramaris@civ.uth.gr

1877-7058 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the organizing committee of the EWaS2 International Conference on Efficient & Sustainable Water Systems Management toward Worth Living Development doi: 10.1016/j.proeng.2016.11.079

1. Introduction

The measurement of the unsteady non uniform flow rate of environmental waters in open channels presents conditions making conventional instruments inappropriate. The most rapid response techniques to measure the velocity distribution and the turbulent characteristics of the flow are an older method (hot-film anemometry) and a modern method (particle image velocimetry).Previous studies report turbulence data obtained from hot-film anemometry ([1], [2], [3], [4]) and from particle image velocimetry ([5], [6], [7]).

Measurements over two different surface geometries (a mesh roughness and spanwise circular rods regularly spaced in the streamwise direction) indicate significant differences in the Reynolds stresses, especially those involving the wall-normal velocity fluctuation, over the outer region. The differences are such that the Reynolds stress anisotropy is smaller over the mesh roughness than the rod roughness. The Reynolds stress anisotropy is largest for a smooth wall. The small-scale anisotropy and intermittency exhibit much smaller differences when the Taylor microscale Reynolds number and the Kolmogorov-normalized mean shear are nominally the same [1].

In the study of [2] results (experimental and computational) for turbulent flow over and within a porous bed have been presented. The simulation has been achieved with the use of rods bundle. The velocity distribution and turbulent characteristics of the flow in the fluid region (over the porous region) was measured with a hot-film anemometry (TSI cross hot film connected with IFA 100). Computed and measured mean velocities and turbulence characteristics indicate the significant influence of the porous bed on the flow characteristics.

In the work of [3] the turbulent flow in open channel with permeable bed is investigated experimentally, with the use of hot film anemometry (TSI cross hot film connected with IFA 100). The permeable bed is simulated initially with porous filters and then with flexible vegetation of porosity s 0.70 and 0.91. The relative porous thickness s'/h (s'=porous thickness, h=flow depth) is varied from 0.375 to 0.500 and the Reynolds number from 7000 to 25000. Measurements of discharge, mean velocity and turbulent characteristics (Reynolds stresses) reveal the effect of the material used (filters, vegetation), s and s'/h on the flow characteristics and the discharge capacity of the channel.In the study [4]the characteristics of turbulent flow above a porous bed are studied experimentally with the use of hotfilm anemometry (TSI cross hot film connected with IFA 100). Velocity and discharge are measured for two different types of porous bed with the same porosity s=0.85 (porous filter and rods bundle) in order to study the distribution of velocity and the turbulence characteristics (Reynolds stresses) of such channels. The experimental results show many differences in both mean and turbulence characteristics between the two bed materials used.

[5] were studied experimentally the characteristics of turbulent flow in an open channel using a particle image velocimetry. Two different types of permeable bed (grass vegetation and gravel bed) with different height (2 and 6cm) with the same porosity (s=0.80) were used to represent the porous bed. Hydraulic characteristics such as distributions of velocities, turbulent intensities and Reynolds stress are investigated and the results show that the bed type can significantly influence the turbulent characteristics of the flow. In the study of [6] the turbulent characteristics of the flow in an open channel with horizontal and inclined impermeable bed were studied experimentally using a particle image velocimetry. Hydraulic characteristics such as distributions of velocities, turbulent intensities and Reynolds stress are investigated at a fine resolution using PIV. Velocity is measured above the horizontal and inclined impermeable bed. The measurements of mean velocity indicate the effect of the channel slope on the flow characteristics. Also, the channel slope influences significantly the turbulent characteristics of the flow such as the variation of longitudinal turbulent intensity the variation of vertical turbulent intensity, the turbulent kinetic energy and Reynolds stresses.

Finally in the study of [7], a series of laboratory experiments were performed to investigate the different impact between rods bundle and other permeable beds. In the other cases the permeable bed is simulated using two different types of vegetation: a) grass vegetation and b) gravel bed. All the permeable bed (rods bundle, grass vegetation and gravel bed) have the same thickness s'=5.5cm and the same porosity s=0.75. The results indicate that the presence of rods bundle impact with different way the velocity distribution in comparison with other permeable beds. Also results showed that the presence of rods bundle in inclined open channels influence in a significantly different way the turbulent characteristics of the flow (turbulent intensities and turbulent shear stresses) in comparison with other permeable beds (grass vegetation or gravel bed).

In this study a comparison between hot-film anemometry and Particle Image Velocimetry (PIV) measurements in open channels is investigated. For this reason experiments with hot-film anemometry method and then with PIV

method in open channels were performed for impermeable and permeable (porous filters) bed. Results show that the different method for experimental measurements influence in some cases the turbulent characteristics of the flow, but not the velocity distribution.

In this study the optimum design of the opening through which a fish refuge pond communicates laterally to the main flow of a river was investigated both experimentally and numerically. That optimum design would be determined by minimizing the potential sedimentation phenomena in the fish refuge basin. The measurements of the flow characteristics in the laboratory were realized with the PIV method, while the mathematical simulations were based on the development and application of a 2D-depth average- hydrodynamic model and a quasi 3D sediment transport model.

2. Experimental Methodology

The experiments with these two methods were conducted in an open channel of the Hydraulics Lab. The channel was 12m long, 25cm wide and 50cm high. Initially, experiments were made with hot-film anemometry method and then with PIV method for two cases of study: a) impermeable bed and b) permeable bed. For the simulation of the permeable bed a porous filter with porous thickness s'=2cm and porosity s=0.75 was used. Measurements of velocity were taken for inclined channel for two different slopes (S=0.002 and S=0.004) and for four different total flow water depths (h= 5cm, 7cm, 9cm and 11cm). Sixteen (16) laboratory experiments were carried out in total, eight (8) for each case. The geometric characteristics of the flow are depicted in figures 1 (for impermeable bed) and 2 (for porous filters). The characteristics of the experiments are presented in Table 1. Reynolds number is greater than 10000 for all cases indicating that the flow is turbulent. The experimental set up is showed in figure 3.

_2_ _V_

Turbulent Flow

ipermeable I

Fig. 1. Geometrical characteristics ofthe flow above impermeable bed

Turbulent Flow

Porous Bed

Impermeable Bed Fig. 2. Geometrical characteristics ofthe flow above porous bed

Table l.Characteristics ofthe Experiments

IMPERMEABLE BED Hot-Film Anemometry Hot-Film Anemometry Particle Image Velocimetry Particle Image Velocimetry

Slope Total Flow Depth h (cm) Umelm Reynolds Number Umelm Reynolds Number

0.002 5 0.201 10050 0.204 10200

0.002 7 0.188 13160 0.189 13230

0.002 9 0.145 13050 0.149 13410

0.002 11 0.110 12100 0.114 12540

0.004 5 0.252 12600 0.256 12800

0.004 7 0.228 15960 0.232 16240

0.004 9 0.191 17190 0.194 17460

0.004 11 0.152 16720 0.155 17050

POROUS FILTERS Hot-Film Hot-Film Particle Image Particle Image

(s'=2cm, e=0.75) Anemometry Anemometry Velocimetry Velocimetry

Slope Flow depth above porous UmeiI Reynolds Number UmeiI Reynolds Number

filters/Total Flow Depth

h' (cm)/ h (cm)

0.002 3/5 0.147 7350 0.154 7700

0.002 5/7 0.113 7910 0.119 8330

0.002 7/9 0.091 8190 0.095 8550

0.002 9/11 0.080 8800 0.087 9570

0.004 3/5 0.168 8400 0.175 8750

0.004 5/7 0.135 9450 0.142 9940

0.004 7/9 0.106 9540 0.115 10350

0.004 9/11 0.094 10340 0.102 11220

Fig. 3. Experimental apparatus

Initially the measurements of velocity and turbulence measurements (horizontal and vertical) were made with the help of a hot film anemometry. For this reason sensors of two components (TSI model 1243) that had the possibility to be placed in the horizontal or vertical level were used. Measurements were taken every 30p.sec. The length of the sensors was 1.02mm and the diameter 50.8p.m. The sensors were placed on a vertical support (TSI probe support) and then they were placed in the channel. For the measurements of shear stress two anemometers, one of older technology (MODEL 1050) and one of newer technology (IFA 100) were used. Small errors (up to 2%) were found due to different anemometers. For the transformation of analogue signal into digital an A/D card was used and data were processed with the help of a program in LABVIEW environment.

Then the measurements of velocity and turbulence measurements (horizontal and vertical) were made using a particle image velocimetry (PIV). The PIV system used for the measurement of the velocity distribution in the flow domain consists of a twin pulsed Nd: Yag lasers (532nm wavelength, 300mjoule/pulse at 10Hz), a cross correlation

8bit IK x IK CCD camera (Kodak MEGAPLUS ES 1.0), a synchronizer, a computer, an image acquisition system and a PIV analysis software (Insight 3G). The laser beams were combined and formed a 1-mm wide sheet by using semi-cylindrical optics. The camera image size creates a 1600 x 1192 pixel array and the dimension of the velocity field was kept to 120 x 110mm for all the experiments. This means that the resolution of the captured images was typically 13.33pixel/mm, whilst the pixel length was 0.075mm/pixel. The laser was installed above the open channel at a distance of 50cm from the water free surface, while the camera viewed from an orthogonal direction.

Twin images were recorded with a temporal interval of 1.5ms. In total, 500 pairs of images were captured per experiment. The temporal interval of these two photographs is 0.25sec.The plane photographs were divided into interrogation spots measuring 16 x 32 pixels (1.2 x 2.4mm). The fluid is generally seeded with tracer particles that, for the purposes of PIV, are generally assumed to follow the flow dynamics [8]. These particles have a size of about 10pm in clean water.The motion of the seeding particles is used to calculate the velocity vectors of the flow. The validation of the images was based on a PIV analysis software (Insight 3G). From the velocity fields, the velocity profiles at various positions were determined using MatLab, which is integrated in the Insight 3G program.

The distance between two neighbour velocity vectors is 1.5mm. From the velocity field we can find the profile of the flow at each vertical direction or the mean space profile in the area. PIVs use the particle concentration method to identify individual particles in an image and follow their flow. The experimental uncertainty of the measured velocity with this technique is approximately ± 2% [9].

The measurements were taken at 8m distance from the entry of channel where the flow is considered fully developed. The full development of flow was checked by comparison of the velocity distributions above the vegetation in two verticals along the flow with 60cm separation distance. The uniformity of the flow was checked with the measurement of flow depth with point gauges at two cross-sections 7m between them. The desirable depth of flow in the downstream section could be adjusted with the help of a weir at the exit of channel. The error of measurement of the flow depth with the point gauge was + 0.1mm. Measurements of velocity were taken at the central axis of the channel cross-section where the flow can be considered that is not influenced by the lateral walls of the channel and depends on the flow depth. The comparison of the velocity distribution at the central axis, with correspondent distributions at symmetrical verticals, 5cm from the central axis showed that the flow is two-dimensional.

3. Analysis of results

The current analysis aims to identify the similarities and differences between hot-film anemometry and particle image velocimetry method on the velocity distribution and turbulent characteristics of the flow. From figures 4 and 5 it is obvious that regarding the velocity distributionthere is a good agreement between these two methods, both for permeable and impermeable bed. The differences in mean velocities and velocity distribution are negligible.

The different method for experimental measurements influence in some cases the turbulent characteristics of the flow. We use dimensionless velocities using the friction velocity U* (U„ = ^gRS, where S is the slope of the channel and R is the hydraulic radius). In figure6 the distribution of turbulence intensity u'/U» within the flow depth (u'/U* vs y/h, with U = VU and u 2 = turbulent normal stress in the flow direction) is presented. In figure 7the distribution of the vertical turbulent intensity v'/U* within the flow depth y/h is shown. In these figures the semi-empirical curve of the distribution of the turbulence intensity in the flow depth for flow above impermeable bed is shown, as given by the relationships of [10].

= 2.30 exp

' £ V h j

■ = 1.27 exp

Regarding the turbulent intensities (longitudinal and vertical turbulent intensity) there are small differences between these two methods (figures 6,7). The most significant difference is for the longitudinal turbulent intensityfor the greater slope (S=0.004) and greater total water flow depth (h=12cm) in which the results of the hot-film anemometry method are greater (approximately until 10-15%) from that from particle image velocimetry method. The

... — . - uv . . .

distribution of turbulent shear stress uv (Reynolds stresses) in the flow depth (-vs y/h) is shown in figure 8.

The greater difference is observed in Reynolds stresses. In all cases (impermeable and permeable bed) the measurements with hot-film anemometry are much higher (over than 30%) in comparison with the measurements from particle image velocimetry. Especially for the porous filters in the case of greater slope (S=0.004) and greater total water flow depth (h=12cm) these is a significant difference between Reynolds stresses (figure 8). This is due to the fact that the hot-film anemometry method has less accuracy and overestimates the Reynolds stresses.

Imperemeable Bed

■ S 0.002 (HOT-FILM)

□ S 0.002 (PIV)

• S 0.004 (HOT-FILM)

O S 0.004 (PIV)

Porous Filters

S=0.002, h=9cm

■ HOT-FILM O PIV

y (cm)

IBBBBB!

U(m/s)

U (m/s)

Fig. 4. Velocity distribution for impermeable bed (h=7cm) Fig. 5. Velocity distribution for porous filters

y/h y/h

Fig. 6. Distribution ofthe longitudinal turbulent intensity in the flow depth y/h

Impermeable Bed

-v'/U*=1.27exp(-y/h)

■ HOT-FILM

• prv

Porous Filters

-v'/U»=1.27exp(-y/h)

■ HOT-FILM • PIV

I 1 I 1 I 1 T

0 0.2 0.4 0.6 0.8

Fig. 7. Distribution ofthe vertical turbulent intensity in the flow depth y/h

0 0.2 0.4 0.6

0 0.2 0.4 0.6

4. Conclusions

Fig. 8. Distribution ofthe turbulent shear stress in the flow depth y/h

In this study a comparison between hot-film anemometry and Particle Image Velocimetry (PIV) measurements in open channels is investigated. For this reason individual experiments with these two methods were conducted. Initially, experiments were made with hot-film anemometry method and then with PIV method in open channels with the same conditions.The follow conclusions can be derived:

The different method for experimental measurements influence in some cases the turbulent characteristics of the flow, especially in the case of permeable bed.

There is a good agreement between these two methods, both for permeable and impermeable bed, regarding the velocity distribution. The differences in mean velocities are negligible.

• On the turbulent intensities (longitudinal and vertical turbulent intensity) there are small differences between these two methods (approximately untillO-15%)

• The greater difference is observed in Reynolds stresses. In all cases (impermeable and permeable bed) the measurements with hot-film anemometry are much higher (over than 30%) in comparison with the measurements from particle image velocimetry. This is due to the fact that the hot-film anemometry method has less accuracy and overestimates the Reynolds stresses.

References

[1] R. Antonia, P. Krogstad, Turbulent structure in boundary layers over different types ofsurface roughness, Fluid Dyn. Res.28 (2001) 139-157.

[2] P. Prinos, D. Sofialidis, E. Keramaris, Turbulent flow over and within a porous bed, J.Hydraul. Res. ASCE 129 (2003) 720-733.

[3] E. Keramaris, P. Prinos, Flow characteristics in open channels with a permeable bed, J. Porous Media 12:2 (2009) 155-165.

[4] E. Keramaris, Comparison between different porous bed (porous filter and rods bundle) in open channels, J. Porous Media 13:2 (2010) 183193.

[5] G. Pechlivanidis, E. Keramaris, I. Pechlivanidis, Experimental study of the effects of permeable bed (grass vegetation and gravel bed) on the turbulentflowusingparticleimagevelocimetry,J.Turbul. 16:1 (2015) 1-16.

[6] E. Keramaris, Effects of inclined impermeable bed on the turbulent characteristics of the flow using particle image velocimetry, J. Turbul. 16:6 (2015) 540-554.

[7] E. Keramaris, G. Pechlivanidis, The different impact of rods bundle in an inclined open channel in comparison with other permeable beds, Proc. of ^^International Conference on Fluid Mechanics and Aerodynamics (FMA 2015), Salerno, Italy, 2015, pp. 119-125.

[8] S.T. Wereley, C.D. Meinhart, Recent advantages in micro-particle image velocimetry, Annu. Rev. Fluid Mech. 42:1 (2010) 557-576.

[9] M. Raffel, C. Willert, S. Wereley, J. Kompenhans, J., Particle Image Velocimetry: A Practical Guide, Springer-Verlag, 2007.

[10] I. Nezu, H. Nakagawa, Turbulence in Open-Channel-Flows, Balkema Publishers, Rotterdam, The Netherlands, 1993.