Unsteady Secondary Motion of Pulsatile Turbulent Flow through a Double 90°-Bend Duct

• Published in: Flow, turbulence and combustion Vol. 104; no. 4; pp. 817 - 833
• Main Authors: Oki, Junichi, Kuga, Yukika, Yamamoto, Ryo, Nakamura, Kazuhiro, Yokohata, Hideaki, Nishida, Keiya, Ogata, Yoichi
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 2020; Springer Nature B.V
• Subjects: Automotive Engineering; Bends; Engineering; Engineering Fluid Dynamics; Engineering Thermodynamics; Fluid dynamics; Fluid flow; Fluid- and Aerodynamics; Heat and Mass Transfer; Particle image velocimetry; Proper Orthogonal Decomposition; Reynolds number; Secondary flow; Switching; Time dependence; Time-frequency analysis; Turbulent flow; Velocity distribution; Wavelet analysis; Wavelet transforms; Swirl switching; Particle image velocimetry; Turbulence; Secondary flow; Pulsatile flow
• ISSN: 1386-6184; 1573-1987
• DOI: 10.1007/s10494-019-00088-y

We investigate turbulent flow with highly pulsating axial velocity passing through a duct with both first and second bends. The time-dependent velocity fields downstream of the bends were measured using time-resolved stereo particle image velocimetry for the steady case (Reynolds number Re  = 36,700) and the pulsatile case ( Re  = 37,800 and Womersley number α  = 59.1). Proper orthogonal decomposition (POD) of the in-plane velocity data isolates the energetic structures of the secondary flow. The modes downstream of the first bend have a Dean motion (mode 0), single swirl (mode 1), and double swirl (mode 2), which agree with those of previous studies on steady turbulent flow. Downstream of the second bend, additional vortices appear in the modes owing to the secondary flow originating in the first bend. The modal structure of the pulsatile case is virtually the same as that of the steady case. To our knowledge, we are the first to find swirl switching in pulsatile flow, whereas the switching has been reported only for steady cases. We further conduct a time-frequency analysis via wavelet transformation onto the POD time coefficient, showing intermittency in energy of the mode associated with swirl switching. Graphical Abstract .