1994) and of the counter-rotating vortex pair (Morton and Ibbetson, 1996, Blanchard et al., 1999). Along with accurate experimental techniques, advanced turbulence models such as direct numerical simulation (DNS) and large eddy simulation (LES) are also used to investigate the flow configuration. Rudman (1996) performed the first direct numerical simulation to predict the time-dependent behavior of a weakly compressible axisymmetric jet exiting normally into a crossflow. Yuan et al. (1999) used LES to study turbulent flow conditions. Eiff et al. (1995) and Eiff and Keffer (1997) used the pattern recognition to study the coherent structures in a jet in crossflow issuing from an elevated exhaust stack. They observed that the wake behind the stack synchronizes with the pseudo-wake of the jet and that some sort of structural connection exists in the far-field between the pseudo-wake and the jet core.
In the present decade (from the year 2000 till date) the number of reported investigations of this flow configuration is quite encouraging. The number of investigations reported in the literature using the advanced experimental techniques such as particle image velocimtery (PIV) and advanced turbulence models such as DNS and LES is continuously increasing with time. Most of the studies deal with the advanced analysis of unsteady flow field, vortex evolution, mixing enhancement and scalar distribution in the flow field. Some important works are reported by Hale et al.
(2000), Hoda and Acharya (2000), Kassimatis et al. (2000), Walter and Leylek (2000), Zhang (2000), Cortelezzi and Karagozian (2001), Keimasi and Rahni (2001), Lim et al. (2001), Rivero et al. (2001), Kalita et al. (2002), Peterson and Plesniak (2002), Johnston et al. (2002), New et al. (2003), Kolar et al. (2003), Naraynan et al.
(2003), Said et al. (2003), Shi et al. (2003), New et al. (2004), Su and Mungal (2004), Sau et al. (2004), Peterson and Plesniak (2004), Wegner et al. (2004), Pathak et al.
(2005), Said et al. (2005), Plesniak and Cusano (2005), Yang and Wang (2005), Mupidi and Maheshan (2005), Guo et al. (2006), Shan and Dimotakis (2006), Jovanovic et al. (2006), Majander and Siikonen (2006), Pathak et al. (2006), Muppidi and Mahesh (2007) and Pathak et al. (2007).
mentioned in Chapter 1 (Fig. 1.2) that four types of vertical structures are formed in the flow field of jets in crossflow. These four structures can be classified broadly into two categories as:
(1) Class 1 structures (2) Class 2 structures
The class 1 structures are formed by the interaction of the jet with the crossflow and the wall and are not observed in free jets. Among structures of this kind, the following three types of vortices are included.
(i) Horseshoe vortices (ii) Wake vortices
(iii)Counter-rotating vortex pair (CRVP)
The class 2 structures are observed in free jets also, but in the flow field of jets in crossflow, their vorticity content, evolution and destabilization are in some way influenced by the presence of the crossflow. Jet-shear-layer vortices are structures of this kind.
2.3.1 Horseshoe Vortices
The formation of a horseshoe-type vortex, around surface mounted obstacles in a uniform stream was reported by several investigators (Baker, 1979, Thomas, 1987, Seal et al., 1995, Sau et al. 2003). The presence of similar horseshoe-type vortices upstream of transverse jets, resulting from the interaction between the jet and the upstream crossflow boundary-layer was experimentally observed by Andreopoulos (1985), Shang et al. (1989), Krothapalli et al. (1990), Fric and Roshko (1994), Kelso and Smits (1995), Morton and Ibbeston (1996), Kelso et al. (1996) and numerically by Kim and Benson (1993), Chiu et al. (1993) and Rudman (1996). All the above investigations were for round jets in crossflow. It was observed that the approaching crossflow wall boundary layer, while encountering an adverse pressure gradient ahead of the jet, separates to form the horseshoe vortices. These vortices wrap around the base of the jet and travel downstream. Kelso et al. (1996) visualized the roll-up mechanism on the upstream side and the vortex breakdown on the downstream side of the horseshoe vortex. Krothapalli et al. (1990), Kelso and Smits (1995), and Rudman (1996) studied how these vortices relate to the shear layer roll-up and the shedding of vortices in the wake of the jet. The horseshoe vortex system can be very complex,
with the number of upstream separation points depending on Reynolds number and the relative magnitudes of the crossflow and jet velocities. The observations made by Kelso and Smits (1995) further reveal that the behaviour of a horseshoe system, can be steady, oscillating, and coalescing, depending on a combination of values of the important flow parameters, such as the Reynolds number and the jet to crossflow velocity ratio. Their results also suggest a strong connection between the unsteadiness in the horseshoe vortex system.
2.3.2 Wake Vortices
The wake vortices are initiated by the entrainment of the crossflow boundary-layer into the wake and the upward re-orientation of the entrained flow into the wake structures. The wake system is the least understood system in the jet in crossflow and has received special attention in many studies. A false or pseudo-wake, with an almost null deficit of the momentum develops downstream of the jet. Experiments and DNS/LES studies of jets in crossflow (McMahon et al., 1971, Fric and Roshko, 1994, Rudman, 1996, Kelso et al., 1996, Morton and Ibbetson 1996, Yuan et al. 1999) have shown that alternate vortices are shed for Reynolds numbers in the range 500 ≤ Re =νua/D ≤ 1000 and jet-to-crossflow velocity ratio R = 3-4, in a way similar to the Karman vortex street formed behind cylinders or solid bluff bodies.
Fric and Roshko (1994) used the smoke-wire technique to visualize the flow field of a round jet in crossflow in a wind tunnel and showed that fluid coming from the wall boundary layer, and shed regularly from the leeward side of the jet, can be detected in the wake as ascending vortices. They found no evidence that the fluid coming from the jet was shed into the wake. Eiff et al. (1995) studied a thermally contaminated round jet issuing from a stack using hot-wire anemometry. The spectral analysis of the temperature signals in the wake showed that the hot fluid was present in the wake with an almost periodic organization. Some contradiction seems to exist between these results and those of Fric and Roshko (1994). Nevertheless, it should be noted that in air the diffusion of smoke (Sc > 1) is smaller than the diffusion of heat, which closely follows the diffusion of vorticity (Pr ≈ 1). The studies of Kelso and Smits (1995) reveal that the origin and formation of the vortices in the wake are fundamentally different from the well-known phenomenon of vortex shedding from
solid bluff bodies. Instead, the wake vortices have their origins in the laminar boundary layer of the wall from which the jet issues.
2.3.3 Counter-Rotating Vortex Pair (CRVP)
The far-field structures of the jet in crossflow that have received most of the attention of researchers are the counter-rotating vortex pair (CRVP). The CRVP are an important feature of the flow, which begin to take form in the near-field of the jet and become dominant in the far-field, where they appear synonymous with the jet. The CRVP is undoubtedly the most studied structure, due to its robustness and distinct appearance in the flow field. All studies seem to indicate that, irrespective of the velocity ratio, the Reynolds number or other parameters (shape of the nozzle, laminar or turbulence boundary-layer, thickness of the boundary-layer, etc.), the counter- rotating pair of vortex is present thus indicating an essential feature of the flow.
However, the instantaneous photographs or numerical simulations of the flow by Smith et al. (1993), Smith and Mungal (1998) and Yuan et al. (1999) show that the symmetrical CRVP is an artefact of the averaging process since the instantaneous velocity field does not contain a symmetrical vortex pair. In addition, Zaman and Foss (1997) and Smith and Mungal (1998) have found that, despite the accurate control of the experimental conditions, sometimes it is difficult to obtain a symmetrical mean velocity field in the far region.
The creation of the counter-rotating vortex pair is still disputed and many hypotheses exist upon the formation of the counter-rotating vortex pair. Notably, a number of experimental and numerical investigations have been carried out to identify the origin of inception of the CRVP. Andreopoulos and Rodi (1984) stated that for velocity ratios (R < 0.5) the CRVP is generated by the interfacial shear between the jet and the crossflow. Fric and Roshko (1994) reported that the CRVP is formed by the shear layer vorticity in the flow field. Kamotani and Greber (1972) suggested that the vortex pair is created in the wake behind the jet, while Crabb et al. (1981) claimed that the CRVP can be visualized upstream of the injection hole and that the pressure gradient in the near hole region is responsible for the formation of CRVP. Broadwell and Breidenthal (1984) compared the formation of the CRVP to the vortex formed by a wing in the air and concluded that a lift force is imparted on the boundary-layer by the penetrating jet forming the CRVP.
Morton and Ibbetson (1996) presented a detailed description of the CRVP production mechanism, suggesting that the puncturing of the wall boundary-layer by the jet creates the CRVP, whose azimuthal vorticity reorients, forming the two vortices.
Yuan et al. (1999) proposed that the CRVP originates from the quasi-steady vortices on each lateral edge of the jet. These vortices encounter an adverse pressure gradient as they approach the lee side of the jet and break down into the CRVP. Walters and Leylek (1997), in a computational study of the streamwise injected jets, concluded that the two mechanisms that produce the CRVP are the interaction of the jet with the cross-stream and in-hole vorticity. They further concluded that the in-hole vorticity was the more significant of the two production mechanisms. Kelso et al. (1996) suggested that the roll-up of the jet-shear-layer is the mechanism that creates the CRVP, while Hale et al. (2000) hypothesized that the CRVP is formed by the roll-up of the boundary-layer fluid outboard of the jet and is intensified in the near jet region by the jet momentum.
Leylek and Zerkle (1994) while simulating the experiments of Pietrzyk et al.
(1988) computed in-hole counter-rotating vortices of the same rotational sense as the primary CRVP in addition to the standard downstream vortex pair. Kohli and Thole (1998) performed calculations of a jet supplied by a plenum and were able to produce in-hole velocity fields containing either a pair of vortices or a single vortex. In both cases, the CRVP developed downstream of the injection, implying that the CRVP is not simply an in-hole vortices advected into the free stream. Furthermore, Lemmon et al. (1999) determined from an Euler inviscid computation that a CRVP forms downstream of an injection hole without any in-hole vorticity supplied and slip along the surface, i.e., no boundary-layer vorticity. The genesis and development of a counter-rotating vortex pair was very well explained by Lim et al. (2001). Their experimental findings indicate that as soon as the developed cylindrical vortex sheet emerges out of the jet pipe, in the presence of the crossflow, it immediately gets folded up on both lateral edges to form the CRVP. Rivero et al. (2001) observed that the CRVP is not a steady feature of the jet in a crossflow, rather the intensity of the two counter-rotating vortices fluctuates strongly with time.
2.3.4 Jet-shear-Layer Vortices
The jet-shear layer vortices are the near-field features of the developed flow field of a jet in a crossflow. The growth of instability-induced leading edge shear layer vortices along the jet/crossflow interface is a very noticeable flow structure, as it continues to dominate the initial portion of the jet. The studies conducted by Fric and Roshko (1994), Kelso et al. (1996), and Lim et al. (2001) for a round jet in crossflow confirm the existence and growth of such shear layer vortices on the initial portion of the jet.
These large-scale structures are either similar to those formed in the shear layer of a round free jet under the Kelvin Helmholtz instability or to the horseshoe vortices formed at a cylinder wall junction. The roll-up process of the shear layer in the jet in crossflow occurs both along the upstream and downstream sides. These vortices are generated due to the Kelvin-Helmholtz like instability of the annular shear layer emanating from the jet orifice (Yuan et al., 1999). On the upstream side the evolution of the shear layer vortices is more evident, clearer and occurs over a larger time scale than that on the downstream side. This is so because in the latter these vortices break down and the mixing processes are very fast, and the roll-up takes place within a couple of diameters.