Lyazid Djenidi | The University of Newcastle (original) (raw)

Papers by Lyazid Djenidi

Physical Review Fluids

The combined effect of wall pulsed jets and roughness in a laminar channel flow is investigated t... more The combined effect of wall pulsed jets and roughness in a laminar channel flow is investigated to determine whether or not turbulence can be triggered and maintained at low Reynolds numbers. The study is carried out through a direct numerical simulation based on the lattice Boltzmann method. The roughness, mounted on both walls of the channel, consists of transverse square bars spanning the whole width of the channel. Rectangular orifice(s) at the bottom wall act as pulsed jets. The jets are pulsed only once with the second jet activated at the end of the cycle of the first jet. Without activating the jets, the flow in the rough wall channel consists mainly in steady secondary motions in the canopies (spaces between bars) and a skimming two-dimensional laminar flow with some oscillations above the canopies. When the jets are activated, a localized three-dimensional turbulence develops and grows in the channel. Not only it is found that this localized turbulence bears strong similarity to that of a fully rough wall turbulent channel flow, but it appears to be in a "pseudo-" fully rough regime, as observed by the relatively negligible (averaged) viscous drag as compared to the form drag generated by the roughness elements. The physical mechanism which allows this to occur is discussed.

Physical Review Fluids

The relation between self-preservation (SP) and the Kolmogorov similarity hypotheses (Kolmogorov,... more The relation between self-preservation (SP) and the Kolmogorov similarity hypotheses (Kolmogorov, The local structure of turbulence in incompressible viscous fluid for very large Reynolds numbers, Dokl. Akad. Nauk SSSR 30, 301 (1941) [Proc. R. Soc. London A 434, 9 (1991]) is investigated through the transport equations for the second-and third-order moments of the longitudinal velocity increments [δu(r,t) = u(x,t) − u(x + r,t), where x, t, and r are the spatial point and the time and longitudinal separation between two points, respectively]. It is shown that the fluid viscosity ν and the mean turbulent kinetic energy dissipation rate (the overbar represents an ensemble average) emerge naturally from the equations of motion as controlling parameters for the velocity increment moments when SP is assumed. Consequently, the Kolmogorov length scale η [≡(ν 3 / ) 1/4 ] and velocity scale v K [≡(ν ) 1/4 ] also emerge as natural scaling parameters in conformity with SP, indicating that Kolmogorov's first hypothesis is subsumed under the more general hypothesis of SP. Further, the requirement for a very large Reynolds number is also relaxed, at least for the first similarity hypothesis. This requirement however is still necessary to derive the two-thirds law (or the four-fifths law) from the analysis. These analytical results are supported by experimental data in wake, jet, and grid turbulence. An expression for the fourth-order moment of the longitudinal velocity increments (δu) 4 is derived from the analysis carried out in the inertial range. The expression, which involves the product of (δu) 2 and ∂δp/∂x, does not require the use the volume-averaged dissipation r , introduced by Oboukhov [Oboukhov, Some specific features of atmospheric turbulence, J. Fluid Mech. 13, 77 (1962)] on a phenomenological basis and used by Kolmogorov to derive his refined similarity hypotheses [Kolmogorov, A refinement of previous hypotheses concerning the local structure of turbulence in a viscous incompressible fluid at high Reynolds number, J. Fluid Mech. 13, 82 (1962)], suggesting that r is not, like , a quantity issuing from the Navier-Stokes equations.

Physical Review Fluids

Hot-wire measurements are carried out in turbulent boundary layers over smooth and rough walls in... more Hot-wire measurements are carried out in turbulent boundary layers over smooth and rough walls in order the assess the behavior of the skewness (S) and flatness (F ) factors of the longitudinal velocity derivative as y, the distance from the wall, increases. The measurements are complemented by direct numerical simulations of a smooth wall turbulent channel flow. It is observed that, as the distance to the wall increases, S and F vary significantly before approaching a constant in the outer layer of the boundary layer. Further, S and F exhibit a nontrivial dependence on the Taylor microscale Reynolds number (Re λ ). For example, in the region below about 0.2δ (δ is the boundary layer thickness) where Re λ varies significantly, S and F strongly vary with Re λ and can be multivalued at a given Re λ . In the outer region, between 0.3δ and 0.6δ, S, F , and Re λ remain approximately constant. The channel flow direct numerical simulation data for S and F exhibit a similar behavior. These results point to the ambiguity that can arise when assessing the Re λ dependence of S and F in wall shear flows. In particular, the multivaluedness of S and F can lead to erroneous conclusions if y/δ is known only poorly, as is the case for the atmospheric shear layer (ASL). If the laboratory turbulent boundary layer is considered an adequate surrogate to the neutral ASL, then the behavior of S and F in the ASL is expected to be similar to that reported here.

Journal of Fluid Mechanics, 2015

A self-preservation (SP) analysis is carried out for a zero pressure gradient (ZPG) rough-wall tu... more A self-preservation (SP) analysis is carried out for a zero pressure gradient (ZPG) rough-wall turbulent boundary layer with a view to establishing the requirements of complete SP (i.e. SP across the entire layer) and determining if these are achievable. The analysis shows that SP is achievable in certain rough-wall boundary layers (irrespectively of the Reynolds number$Re$), when the mean viscous stress is zero or negligible compared to the form drag across the entire boundary layer. In this case, the velocity scale$u^{\ast }$must be constant, the length scale$l$should vary linearly with the streamwise distance$x$and the roughness height$k$must be proportional to$l$. Although this result is consistent with that of Rotta (Prog. Aeronaut. Sci., vol. 2 (1), 1962, pp. 1–95), it is derived in a more rigorous manner than the method employed by Rotta. Further, it is noted that complete SP is not possible in a smooth-wall ZPG turbulent boundary layer. The SP conditions are tested against p...

Springer Proceedings in Physics, 2016

The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) ... more The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) 2 ae are used as a scale-by-scale budget to quantify the effect of initial conditions at low Reynolds numbers, typical of grid turbulence. The validity of these equations is first investigated via hot-wire measurements of velocity and transverse vorticity fluctuations. The transport equation for AE(dq) 2 ae is shown to be balanced at all scales, while anisotropy of the large scales leads to a significant imbalance in the equation for AE(du) 2 ae. The effect of using similarity to evaluate the transport equation is rigorously tested. This approach has the desirable benefit of requiring less extensive measurements to calculate the inhomogeneous term of the transport equation. The similarity form of the AE(dq) 2 ae equation produces nearly identical results as those obtained without the similarity assumption. In the case of the AE(du) 2 ae equation, the similarity method forces a balance at large separation, although the imbalance due to large scale anisotropy remains. The initial conditions of the turbulence at constant R M . 10,400 (28 £ R k £ 55) are changed by using three grids of different geometries. Initial conditions affect the shape and magnitude of the second-and third-order structure functions, as well as the anisotropy of the large scales. The effect of initial conditions on the scale-by-scale budget is restricted to the inhomogeneous term of the transport equations, while the dissipation term remains unaffected despite the low R k . Scales as small as k are affected by the changes in initial conditions.

International Journal of Heat and Fluid Flow, Apr 30, 1997

Statistics of #u/Oy haw,= been measured in a turbulent boundary layer over either a smooth wall o... more Statistics of #u/Oy haw,= been measured in a turbulent boundary layer over either a smooth wall or a riblet surface using parallel hot wires. Over the riblets, the mean square value of Ou/#y (normalised with smooth wall variables) is smaller than over the smooth wall; the difference increases as the distance to the wall decreases. Also this difference is larger when the riblets operate in a drag-augmenting mode. These changes would reflect changes in the near-wall turbulence structure. Furthermore, on the basis that Ou/#y is a major contributor to the spanwise vorticity, the results indicate that, relative to the smooth wall, the mean square spanwise vortieity is reduced near a riblet surface. © 1997 by Elsevier Science Inc.

Decaying grid turbulence is considered at low Reynolds number (R λ ~ 50) for different initial co... more Decaying grid turbulence is considered at low Reynolds number (R λ ~ 50) for different initial conditions. Three different grid geometries are used. Heat is injected via a mandoline at a distance of 1.5 M from the grid. The amount of heating is such that temperature may be treated as a passive scalar. A small contraction (1.36:1) is added at a distance of 11M downstream of the grid. The power-law exponents for the scalar variance are compared with those for the turbulent kinetic energy. These exponents depend on the grid geometry.

International Journal of Heat and Fluid Flow, Jun 1, 2004

The organised motion in a turbulent channel flow with a succession of square bars on the bottom w... more The organised motion in a turbulent channel flow with a succession of square bars on the bottom wall has been investigated using direct numerical simulations. Several values of the ratio w/k, where k is the bar height and w is the longitudinal separation between consecutive bars have been examined in detail. Relative to a smooth surface, the streamwise extent of the near-wall structures is decreased while their spanwise extent is increased. As w/k increases, the coherence decreases in the streamwise direction having a minimum for w/k=7. This is due to the outward motion occurring, most of all, near the leading edge of the elements. The Reynolds stress anisotropy tensor and its invariants show a closer approach to isotropy over the rough wall than over a smooth wall.

Physics of Fluids, Jul 1, 1994

ABSTRACT

Experiments in Fluids, 2010

Measurements downstream of a grid followed by a contraction with an area ratio of 1.36 have been ... more Measurements downstream of a grid followed by a contraction with an area ratio of 1.36 have been made with a one-component vorticity probe. The velocity data indicate that there is indeed an improvement in isotropy of the large scales when such a contraction is used. In particular, the spectrum of u loses a significant amount of low frequency energy and its peak is shifted towards that of v. Although the v spectrum is less affected than the u spectrum, the relative magnitudes of the u and v spectra satisfy isotropy more closely with than without the contraction. In particular, isotropy is most closely approximated when the contraction is used in combination with a grid designed to minimise vortex shedding. The lateral vorticity spectrum is essentially unaffected, irrespective of the grid that is used or the magnitude of the wavenumber.

Experiments in Fluids, 2005

The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) ... more The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) 2 ae are used as a scale-by-scale budget to quantify the effect of initial conditions at low Reynolds numbers, typical of grid turbulence. The validity of these equations is first investigated via hot-wire measurements of velocity and transverse vorticity fluctuations. The transport equation for AE(dq) 2 ae is shown to be balanced at all scales, while anisotropy of the large scales leads to a significant imbalance in the equation for AE(du) 2 ae. The effect of using similarity to evaluate the transport equation is rigorously tested. This approach has the desirable benefit of requiring less extensive measurements to calculate the inhomogeneous term of the transport equation. The similarity form of the AE(dq) 2 ae equation produces nearly identical results as those obtained without the similarity assumption. In the case of the AE(du) 2 ae equation, the similarity method forces a balance at large separation, although the imbalance due to large scale anisotropy remains. The initial conditions of the turbulence at constant R M . 10,400 (28 £ R k £ 55) are changed by using three grids of different geometries. Initial conditions affect the shape and magnitude of the second-and third-order structure functions, as well as the anisotropy of the large scales. The effect of initial conditions on the scale-by-scale budget is restricted to the inhomogeneous term of the transport equations, while the dissipation term remains unaffected despite the low R k . Scales as small as k are affected by the changes in initial conditions.

Exp Fluid, 2008

Particle Image Velocimetry (PIV) measurements and Planar Laser Induced Fluorescence (PLIF) visual... more Particle Image Velocimetry (PIV) measurements and Planar Laser Induced Fluorescence (PLIF) visualizations have been made in a turbulent boundary layer over a rough wall. The roughness consists of square bars placed transversely to the flow at a pitch to height ratio of 11. The near wall region flow is dominated by coherent structures associated with the shear layers which originate at the downstream edge of a roughness element and reattach on the wall upstream of the subsequent element. Shear layer vortices shed downstream a roughness elements impact on the consecutive element producing intermittent form drag and strong turbulence production.

The effect of two different types of surface roughness on a turbulent boundary layer was studied ... more The effect of two different types of surface roughness on a turbulent boundary layer was studied using 2-component LDV measurements in a relatively high speed water tunnel. One roughness consists of square bars at a streamwise spacing p equal to 2k (k is the roughness height). The other consists of cylindrical rods with p/k equal to 4. Both roughnesses are aligned in a direction transverse to the flow. Measurements of the turbulent field were carried out over a wide range of Reynolds numbers, (1 500 < R θ < 23 000) based on the momentum thickness. Comparison of the turbulent field between different surfaces is made at R θ ∼ 9 000. This study supports previous attempts to classify rough surfaces according to their turbulence characteristics, and extends them by providing measurements at high Reynolds numbers for three distinct surface conditions.

The present paper reports Laser Doppler Velocimetry (LDV) measurements, and Planar Laser Induced ... more The present paper reports Laser Doppler Velocimetry (LDV) measurements, and Planar Laser Induced Fluorescence (PLIF) visualisations in a turbulent boundary layer subjected to suction applied through a porous strip. The focus of the study is to investigate the effect suction can have on the boundary layer in the near-wall region, downstream of the suction strip. The mean velocity and fluctuation distributions indicate that when suction is relatively high, pseudo-relaminarisation takes place behind the suction strip. The pseudo-relaminaristion, although not complete, is accompanied by a change in the near-wall structure of the boundary layer. This change is marked by the effect suction has on the low speed streaks, which are associated with the near-wall quasi-streamwise vortices. When suction is applied at relatively high suction rates, the low-speed streaks are less agitated and remain closer to the wall over a longer distance in comparison to what is observed when there is no suction. The effect of suction of the near-wall structure is also reflected in the space correlation coefficient in the spanwise direction. When suction is applied, the correlation coefficient is higher than that when there is no suction, suggesting that the spanwise coherence of the near-wall structure increases. This would suggest that suction would tend to reduce the threedimensionality features of the near-wall structure of the layer. A procedure for detecting the low-speed streaks has being developed. The procedure is based on reading an image, filtering it, detecting the edges of streaks and removing the background noise. The procedure is currently tested. Preliminary results indicate that this technique can detect quite reliably the presence of streaks, despite some minor difficulties. Work is being carried out to improve the streak detection. Once the procedure is fully operational and calibrated it will be used to quantify the effect of suction on the streaks, and particularly on their spanwise spacing and the frequency of their bursting.

Physical Review Fluids

The combined effect of wall pulsed jets and roughness in a laminar channel flow is investigated t... more The combined effect of wall pulsed jets and roughness in a laminar channel flow is investigated to determine whether or not turbulence can be triggered and maintained at low Reynolds numbers. The study is carried out through a direct numerical simulation based on the lattice Boltzmann method. The roughness, mounted on both walls of the channel, consists of transverse square bars spanning the whole width of the channel. Rectangular orifice(s) at the bottom wall act as pulsed jets. The jets are pulsed only once with the second jet activated at the end of the cycle of the first jet. Without activating the jets, the flow in the rough wall channel consists mainly in steady secondary motions in the canopies (spaces between bars) and a skimming two-dimensional laminar flow with some oscillations above the canopies. When the jets are activated, a localized three-dimensional turbulence develops and grows in the channel. Not only it is found that this localized turbulence bears strong similarity to that of a fully rough wall turbulent channel flow, but it appears to be in a "pseudo-" fully rough regime, as observed by the relatively negligible (averaged) viscous drag as compared to the form drag generated by the roughness elements. The physical mechanism which allows this to occur is discussed.

Physical Review Fluids

The relation between self-preservation (SP) and the Kolmogorov similarity hypotheses (Kolmogorov,... more The relation between self-preservation (SP) and the Kolmogorov similarity hypotheses (Kolmogorov, The local structure of turbulence in incompressible viscous fluid for very large Reynolds numbers, Dokl. Akad. Nauk SSSR 30, 301 (1941) [Proc. R. Soc. London A 434, 9 (1991]) is investigated through the transport equations for the second-and third-order moments of the longitudinal velocity increments [δu(r,t) = u(x,t) − u(x + r,t), where x, t, and r are the spatial point and the time and longitudinal separation between two points, respectively]. It is shown that the fluid viscosity ν and the mean turbulent kinetic energy dissipation rate (the overbar represents an ensemble average) emerge naturally from the equations of motion as controlling parameters for the velocity increment moments when SP is assumed. Consequently, the Kolmogorov length scale η [≡(ν 3 / ) 1/4 ] and velocity scale v K [≡(ν ) 1/4 ] also emerge as natural scaling parameters in conformity with SP, indicating that Kolmogorov's first hypothesis is subsumed under the more general hypothesis of SP. Further, the requirement for a very large Reynolds number is also relaxed, at least for the first similarity hypothesis. This requirement however is still necessary to derive the two-thirds law (or the four-fifths law) from the analysis. These analytical results are supported by experimental data in wake, jet, and grid turbulence. An expression for the fourth-order moment of the longitudinal velocity increments (δu) 4 is derived from the analysis carried out in the inertial range. The expression, which involves the product of (δu) 2 and ∂δp/∂x, does not require the use the volume-averaged dissipation r , introduced by Oboukhov [Oboukhov, Some specific features of atmospheric turbulence, J. Fluid Mech. 13, 77 (1962)] on a phenomenological basis and used by Kolmogorov to derive his refined similarity hypotheses [Kolmogorov, A refinement of previous hypotheses concerning the local structure of turbulence in a viscous incompressible fluid at high Reynolds number, J. Fluid Mech. 13, 82 (1962)], suggesting that r is not, like , a quantity issuing from the Navier-Stokes equations.

Physical Review Fluids

Hot-wire measurements are carried out in turbulent boundary layers over smooth and rough walls in... more Hot-wire measurements are carried out in turbulent boundary layers over smooth and rough walls in order the assess the behavior of the skewness (S) and flatness (F ) factors of the longitudinal velocity derivative as y, the distance from the wall, increases. The measurements are complemented by direct numerical simulations of a smooth wall turbulent channel flow. It is observed that, as the distance to the wall increases, S and F vary significantly before approaching a constant in the outer layer of the boundary layer. Further, S and F exhibit a nontrivial dependence on the Taylor microscale Reynolds number (Re λ ). For example, in the region below about 0.2δ (δ is the boundary layer thickness) where Re λ varies significantly, S and F strongly vary with Re λ and can be multivalued at a given Re λ . In the outer region, between 0.3δ and 0.6δ, S, F , and Re λ remain approximately constant. The channel flow direct numerical simulation data for S and F exhibit a similar behavior. These results point to the ambiguity that can arise when assessing the Re λ dependence of S and F in wall shear flows. In particular, the multivaluedness of S and F can lead to erroneous conclusions if y/δ is known only poorly, as is the case for the atmospheric shear layer (ASL). If the laboratory turbulent boundary layer is considered an adequate surrogate to the neutral ASL, then the behavior of S and F in the ASL is expected to be similar to that reported here.

Journal of Fluid Mechanics, 2015

A self-preservation (SP) analysis is carried out for a zero pressure gradient (ZPG) rough-wall tu... more A self-preservation (SP) analysis is carried out for a zero pressure gradient (ZPG) rough-wall turbulent boundary layer with a view to establishing the requirements of complete SP (i.e. SP across the entire layer) and determining if these are achievable. The analysis shows that SP is achievable in certain rough-wall boundary layers (irrespectively of the Reynolds number$Re$), when the mean viscous stress is zero or negligible compared to the form drag across the entire boundary layer. In this case, the velocity scale$u^{\ast }$must be constant, the length scale$l$should vary linearly with the streamwise distance$x$and the roughness height$k$must be proportional to$l$. Although this result is consistent with that of Rotta (Prog. Aeronaut. Sci., vol. 2 (1), 1962, pp. 1–95), it is derived in a more rigorous manner than the method employed by Rotta. Further, it is noted that complete SP is not possible in a smooth-wall ZPG turbulent boundary layer. The SP conditions are tested against p...

Springer Proceedings in Physics, 2016

The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) ... more The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) 2 ae are used as a scale-by-scale budget to quantify the effect of initial conditions at low Reynolds numbers, typical of grid turbulence. The validity of these equations is first investigated via hot-wire measurements of velocity and transverse vorticity fluctuations. The transport equation for AE(dq) 2 ae is shown to be balanced at all scales, while anisotropy of the large scales leads to a significant imbalance in the equation for AE(du) 2 ae. The effect of using similarity to evaluate the transport equation is rigorously tested. This approach has the desirable benefit of requiring less extensive measurements to calculate the inhomogeneous term of the transport equation. The similarity form of the AE(dq) 2 ae equation produces nearly identical results as those obtained without the similarity assumption. In the case of the AE(du) 2 ae equation, the similarity method forces a balance at large separation, although the imbalance due to large scale anisotropy remains. The initial conditions of the turbulence at constant R M . 10,400 (28 £ R k £ 55) are changed by using three grids of different geometries. Initial conditions affect the shape and magnitude of the second-and third-order structure functions, as well as the anisotropy of the large scales. The effect of initial conditions on the scale-by-scale budget is restricted to the inhomogeneous term of the transport equations, while the dissipation term remains unaffected despite the low R k . Scales as small as k are affected by the changes in initial conditions.

International Journal of Heat and Fluid Flow, Apr 30, 1997

Statistics of #u/Oy haw,= been measured in a turbulent boundary layer over either a smooth wall o... more Statistics of #u/Oy haw,= been measured in a turbulent boundary layer over either a smooth wall or a riblet surface using parallel hot wires. Over the riblets, the mean square value of Ou/#y (normalised with smooth wall variables) is smaller than over the smooth wall; the difference increases as the distance to the wall decreases. Also this difference is larger when the riblets operate in a drag-augmenting mode. These changes would reflect changes in the near-wall turbulence structure. Furthermore, on the basis that Ou/#y is a major contributor to the spanwise vorticity, the results indicate that, relative to the smooth wall, the mean square spanwise vortieity is reduced near a riblet surface. © 1997 by Elsevier Science Inc.

Decaying grid turbulence is considered at low Reynolds number (R λ ~ 50) for different initial co... more Decaying grid turbulence is considered at low Reynolds number (R λ ~ 50) for different initial conditions. Three different grid geometries are used. Heat is injected via a mandoline at a distance of 1.5 M from the grid. The amount of heating is such that temperature may be treated as a passive scalar. A small contraction (1.36:1) is added at a distance of 11M downstream of the grid. The power-law exponents for the scalar variance are compared with those for the turbulent kinetic energy. These exponents depend on the grid geometry.

International Journal of Heat and Fluid Flow, Jun 1, 2004

The organised motion in a turbulent channel flow with a succession of square bars on the bottom w... more The organised motion in a turbulent channel flow with a succession of square bars on the bottom wall has been investigated using direct numerical simulations. Several values of the ratio w/k, where k is the bar height and w is the longitudinal separation between consecutive bars have been examined in detail. Relative to a smooth surface, the streamwise extent of the near-wall structures is decreased while their spanwise extent is increased. As w/k increases, the coherence decreases in the streamwise direction having a minimum for w/k=7. This is due to the outward motion occurring, most of all, near the leading edge of the elements. The Reynolds stress anisotropy tensor and its invariants show a closer approach to isotropy over the rough wall than over a smooth wall.

Physics of Fluids, Jul 1, 1994

ABSTRACT

Experiments in Fluids, 2010

Measurements downstream of a grid followed by a contraction with an area ratio of 1.36 have been ... more Measurements downstream of a grid followed by a contraction with an area ratio of 1.36 have been made with a one-component vorticity probe. The velocity data indicate that there is indeed an improvement in isotropy of the large scales when such a contraction is used. In particular, the spectrum of u loses a significant amount of low frequency energy and its peak is shifted towards that of v. Although the v spectrum is less affected than the u spectrum, the relative magnitudes of the u and v spectra satisfy isotropy more closely with than without the contraction. In particular, isotropy is most closely approximated when the contraction is used in combination with a grid designed to minimise vortex shedding. The lateral vorticity spectrum is essentially unaffected, irrespective of the grid that is used or the magnitude of the wavenumber.

Experiments in Fluids, 2005

The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) ... more The transport equations for the second-order velocity structure functions AE(du) 2 ae and AE(dq) 2 ae are used as a scale-by-scale budget to quantify the effect of initial conditions at low Reynolds numbers, typical of grid turbulence. The validity of these equations is first investigated via hot-wire measurements of velocity and transverse vorticity fluctuations. The transport equation for AE(dq) 2 ae is shown to be balanced at all scales, while anisotropy of the large scales leads to a significant imbalance in the equation for AE(du) 2 ae. The effect of using similarity to evaluate the transport equation is rigorously tested. This approach has the desirable benefit of requiring less extensive measurements to calculate the inhomogeneous term of the transport equation. The similarity form of the AE(dq) 2 ae equation produces nearly identical results as those obtained without the similarity assumption. In the case of the AE(du) 2 ae equation, the similarity method forces a balance at large separation, although the imbalance due to large scale anisotropy remains. The initial conditions of the turbulence at constant R M . 10,400 (28 £ R k £ 55) are changed by using three grids of different geometries. Initial conditions affect the shape and magnitude of the second-and third-order structure functions, as well as the anisotropy of the large scales. The effect of initial conditions on the scale-by-scale budget is restricted to the inhomogeneous term of the transport equations, while the dissipation term remains unaffected despite the low R k . Scales as small as k are affected by the changes in initial conditions.

Exp Fluid, 2008

Particle Image Velocimetry (PIV) measurements and Planar Laser Induced Fluorescence (PLIF) visual... more Particle Image Velocimetry (PIV) measurements and Planar Laser Induced Fluorescence (PLIF) visualizations have been made in a turbulent boundary layer over a rough wall. The roughness consists of square bars placed transversely to the flow at a pitch to height ratio of 11. The near wall region flow is dominated by coherent structures associated with the shear layers which originate at the downstream edge of a roughness element and reattach on the wall upstream of the subsequent element. Shear layer vortices shed downstream a roughness elements impact on the consecutive element producing intermittent form drag and strong turbulence production.

The effect of two different types of surface roughness on a turbulent boundary layer was studied ... more The effect of two different types of surface roughness on a turbulent boundary layer was studied using 2-component LDV measurements in a relatively high speed water tunnel. One roughness consists of square bars at a streamwise spacing p equal to 2k (k is the roughness height). The other consists of cylindrical rods with p/k equal to 4. Both roughnesses are aligned in a direction transverse to the flow. Measurements of the turbulent field were carried out over a wide range of Reynolds numbers, (1 500 < R θ < 23 000) based on the momentum thickness. Comparison of the turbulent field between different surfaces is made at R θ ∼ 9 000. This study supports previous attempts to classify rough surfaces according to their turbulence characteristics, and extends them by providing measurements at high Reynolds numbers for three distinct surface conditions.

The present paper reports Laser Doppler Velocimetry (LDV) measurements, and Planar Laser Induced ... more The present paper reports Laser Doppler Velocimetry (LDV) measurements, and Planar Laser Induced Fluorescence (PLIF) visualisations in a turbulent boundary layer subjected to suction applied through a porous strip. The focus of the study is to investigate the effect suction can have on the boundary layer in the near-wall region, downstream of the suction strip. The mean velocity and fluctuation distributions indicate that when suction is relatively high, pseudo-relaminarisation takes place behind the suction strip. The pseudo-relaminaristion, although not complete, is accompanied by a change in the near-wall structure of the boundary layer. This change is marked by the effect suction has on the low speed streaks, which are associated with the near-wall quasi-streamwise vortices. When suction is applied at relatively high suction rates, the low-speed streaks are less agitated and remain closer to the wall over a longer distance in comparison to what is observed when there is no suction. The effect of suction of the near-wall structure is also reflected in the space correlation coefficient in the spanwise direction. When suction is applied, the correlation coefficient is higher than that when there is no suction, suggesting that the spanwise coherence of the near-wall structure increases. This would suggest that suction would tend to reduce the threedimensionality features of the near-wall structure of the layer. A procedure for detecting the low-speed streaks has being developed. The procedure is based on reading an image, filtering it, detecting the edges of streaks and removing the background noise. The procedure is currently tested. Preliminary results indicate that this technique can detect quite reliably the presence of streaks, despite some minor difficulties. Work is being carried out to improve the streak detection. Once the procedure is fully operational and calibrated it will be used to quantify the effect of suction on the streaks, and particularly on their spanwise spacing and the frequency of their bursting.