Experimental investigation on streamlines in a 180o sharp bend

One of the most important concerns of hydraulics engineers is predicting erosion at the outer banks of rivers by studying the flow pattern along the bend. Not only are the streamlines in meanders non parallel curved lines, but they are also twisted. To study the streamlines flowing a sharp bend, a 180o sharp bend was constructed at Persian Gulf University in Iran. Three dimensional flow components at different locations of the bend were measured using Vectrino velocim. In this paper, streamlines were drawn and investigated in different longitudinal profiles, cross sections, and plans. The results indicated that the secondary flow strength and size of the vortex formed at the distance of the beginning to the bend apex would increase. The core of central vortex moved away from the inner bank towards central line of the channel by 22%, and to the water surface by 20%. On the contrary, the size of the secondary vortex increased by 15%. In addition, the average of the horizontal angle of the streamlines, vector and the locus of maximum velocity were determined at different levels in the present investigation.


Introduction
The secondary flow is formed as a result of centrifugal force and its interaction with lateral pressure gradients due to lateral slope of water surface.In this flow, water moves away at the upper part of the river, and at the lower part, it moves towards the inner bank.In open-channel bends, the curvature of the flow gives rise to secondary flow, resulting in the helical motion.This helical motion is of high importance in meandering rivers, where it plays a key role in erosion and sedimentation patterns in the river's bed (Van Balen, Uijttewaal, & Blanckaert, 2009).Therefore, it is vital to know about and study flow pattern in river bends in order to predict and prevent outer wall erosion in the rivers.Moreover, the proper understanding of flow characteristics in curved open-channels is vital in predicting the spreading of pollutants and thus for water quality of natural river systems (Van Balen et al., 2009).
Since scour and flow patterns are of high importance in river and hydraulics engineering, a great number of researchers have always conducted studies on flow structure and sediment transport through bend and straight reaches.Kra and Merkley (2004) developed a computational method based on mathematical modeling for both two-dimensional and three-dimensional velocity distributions for steady-state uniform flow in open channels of rectangular cross-section.It is evident that the twodimensional version of the model is not appropriate for calculating surface velocity coefficients.Sui, Fang, and Karney (2006) carried out an experimental study on local scour in a flume with a 90º bend and analyzed the effect of some param including the Froude number, slope and width of the protective wall, and size of bed particle on the scour at bed level.Huang, Jia, Hsun-Chuan, and Sam (2009) applied NCCHE3D 3-D free surface to study secondary flows in an experimental channel.The agreements of vertically-averaged velocities between the simulated results were obtained by using different turbulence models with different pressure solution techniques, and the resulting measured data were satisfactory.Wang, Zhou, and Shao (2010) used a computational fluid dynamics model for simulation of two-dimensional water flow, sediment transport, bank failure processes, and the subsequent channel pattern changes.They considered the effects of secondary currents in bend channel and validated the water flow model using experimental data.Experimental and numerical studies of flow pattern in a 90º bend by Abhari, Ghodsian, Vaghefi, and Panahpur (2010) indicated that streamlines at the level close to the bed orient to the inner wall and at levels near water surface decline to the outer wall.Chan, Zhang, Leu, and Chen (2010) studied the turbulent flow in a channel with periodic porous ribs on one wall.They used the Reynolds averaged Navier-Stokes (Rans) equations with a k-ɛ turbulent model for turbulence closure.Barbhuiya and Talukdar (2010) carried out an experimental study of three dimensional flow and scour pattern in a 90º bend, and measured the time averaged velocity components, turbulent intensity components and Reynolds stresses in different vertical sections by using ADV.The Results showed that the maximum measured velocity is 1.61 times the mean velocity.Stoesser, Ruether, and Olsen (2010) solved the Navier-Stokes equations on a fine threedimensional grid by using a Large Eddy Simulation approach and a method that is based on the Rans equations for which there are two different isotropic turbulence closures.The results provided clear evidence that the Rans code was able to predict the time-averaged primary velocities with good agreement regardless of the turbulence model used.Bonakdari, Baghalian, Nazari, and Fazli (2011) predicted flow field in a mild 90º bend using Artificial Neural Networks (ANN) and Genetic Algorithm (GA).They studied the variations of velocities in both experimental and numerical (CFD) models.Moreover, they compared the results of ANN and CFX methods in sections where experimental data were not available.Constantinescu, Koken, and Zeng (2011) considered the flow in an open channel bend of strong curvature (the ratio between the radius of curvature of the curved reach and the channel width is close to 1.3) over realistic topography corresponding to equilibrium scour conditions.Results demonstrated that compared to Rans, DES (detached eddy simulation) is able to better capture the redistribution of the mean flow stream wise velocity.Baghalian, Bonakdari, Nazari, and Fazli (2012) investigated the velocity field in a 90 degree open channel bend using artificial intelligence, analytical, experimental, and numerical methods.They indicated that numerical, ANN and experimental results could show that the maximum velocity occurs under the free-surface but the analytical solution could not.Blanckaert et al. (2013) studied three distinct processes of flow separation near the banks in sharply-curved open-channel bends.The experiments were performed with both a flat immobile gravel bed and a mobile sand bed with dominant bed load sediment transport.Gholami, Akhtari, Minatour, Bonakdari, and Javadi (2014) carried out the experimental and numerical modelling of flow pattern at a strongly-curved 90 degree bend and reported that in both models, along the bend, the maximum velocity always occurs near the inner wall while the minimum occurs near the outer wall.Celik, Diplas, and Dancey (2014) measured the pressure fluctuations on the surface of a coarse, fully exposed, spherical grain resting upon a bed of identical grains in an open channel turbulent flow.They concluded that the stream wise velocity near the bed is most directly related to those force events crucial to particle entrainment.Huang, Li, Huang, and Liou (2014) acquired temperature profiles in a PDMS micro channel with a 90º sharp bend using a molecule-based temperature sensor.These temperature evolutions agree with secondary flow patterns identified from the velocity measurement.Vaghefi, Akbari, and Fiouz (2014) used the Depth-Averaged method to study and analyze shear stress distribution near the bed in a 180º sharp bend flume.The results suggested that the maximum dimensionless shear stress occurs near the inner wall and at the 40º cross section.Vaghefi, Akbari, and Fiouz (2015) measured three dimensional flow velocity components in a 180 degree sharp bend.The comparison between the longitudinal velocity values at distances of 5 and 95% from the bed showed a 60% increase in flow velocity from near the bed to water surface.Horvat, Isic, and Spasojevic (2015)

Results and di
In this se affecting the bend, the st profile, cross s Then, the ave determined a maximum v addressed.

Streamlines in c
When flow occur in its th due to ex and longitudin m.Technology Figure 2. The lo bend for collectin Figure 4. Strea 80, d) 90, e) 10 As seen clockwise v secondary v Maringá, v. 39, n.
figures, there section, is formed at Figure degree and f) Stream In differ and n the fl can b the in that outer phen bend away sharp the d featu longi stream concl layer bed i being midstream the Figur secon strike stream direc

Figure
Figure 6.Stream distances of a) 5, b The streamlines In Figure 7 of the stream been compare m.Technology Figure 8 of 180 d of flow d Firs the lon that it keeps t all dire Figure toward gradien depth o of seco occurs the ben surface oriente beginni