Fuqi Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Fuqi Li has a total of 39 co-authors. Here are the top co-authors by number of shared publications:
- Peng Wang
Shanghai Jiao Tong University
8 shared publications - Yingzheng Liu
Shanghai Jiao Tong University
5 shared publications - Yuzheng Li
School of Mechanical Engineering and Automation, Fuzhou University, Fuzhou, China
4 shared publications Bang Liu
2 shared publications- Yinqi Wu2 shared publications
- Bang Liu1 shared publication
Latest Publications
- 27 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-13
RBFNN-Assisted Decoupling Direct/Indirect Adaptive Robust Control for Pump-Valve Coordinated Steering System
Traditional electro-hydraulic steering system (EHSS) in heavy-duty multiaxle vehicle suffers from high energy consumption, limited steering modes, and inadequate accuracy. To solve these challenges, a pump-valve coordinated (PVC) EHSS has been gradually proposed. However, existing control methods for this system still exhibit insufficient tracking accuracy, slow error convergence rate, and compromised safety under demanding conditions. Therefore, this article proposes a radial basis function neural network (RBFNN)-assisted decoupling direct/indirect adaptive robust controller. The direct/indirect adaptive robust control (DIARC) framework is introduced to effectively handle high-order nonlinearities of the system and time-varying damping parameters of the hydraulic cylinders. A decoupling strategy is proposed to address flow coupling between the pump and valve subsystems, ensuring precise flow distribution. Furthermore, an RBFNN is integrated to rapidly predict and compensate for unmodeled mechanical dynamics and external disturbances. Theoretically, the proposed control method achieves a comprehensive compensation for both parametric uncertainties and highly nonlinear unknown dynamics. Meanwhile comparative experiments demonstrate that the proposed controller substantially outperforms traditional methods under various steering conditions, particularly in high-frequency micro-adjustment scenarios, where it effectively suppresses mechanical high-frequency disturbances, achieves rapid error convergence, and delivers the best tracking performance.
- 27 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-13
RBFNN-Assisted Decoupling Direct/Indirect Adaptive Robust Control for Pump-Valve Coordinated Steering System
Traditional electro-hydraulic steering system (EHSS) in heavy-duty multiaxle vehicle suffers from high energy consumption, limited steering modes, and inadequate accuracy. To solve these challenges, a pump-valve coordinated (PVC) EHSS has been gradually proposed. However, existing control methods for this system still exhibit insufficient tracking accuracy, slow error convergence rate, and compromised safety under demanding conditions. Therefore, this article proposes a radial basis function neural network (RBFNN)-assisted decoupling direct/indirect adaptive robust controller. The direct/indirect adaptive robust control (DIARC) framework is introduced to effectively handle high-order nonlinearities of the system and time-varying damping parameters of the hydraulic cylinders. A decoupling strategy is proposed to address flow coupling between the pump and valve subsystems, ensuring precise flow distribution. Furthermore, an RBFNN is integrated to rapidly predict and compensate for unmodeled mechanical dynamics and external disturbances. Theoretically, the proposed control method achieves a comprehensive compensation for both parametric uncertainties and highly nonlinear unknown dynamics. Meanwhile comparative experiments demonstrate that the proposed controller substantially outperforms traditional methods under various steering conditions, particularly in high-frequency micro-adjustment scenarios, where it effectively suppresses mechanical high-frequency disturbances, achieves rapid error convergence, and delivers the best tracking performance.
- 1 February 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38 (2)
Delayed detached-eddy simulation in ducted orifices jet dynamics for low-pressure compressed air energy storage system
In this study, turbulent flow through an orifice was numerically modeled using delayed detached-eddy simulations to clarify transient flow dynamics and coherent flow structures. Three inlet pressures (Pin = 1 bar, 2 bars, and 3 bars) were selected for comparative analysis under a typical orifice porosity (orifice hole area to pipe area ratio) of β = 28.4%. The numerical results agreed well with the previous experimental measurements. The unsteady jet characteristics were first investigated regarding the time-averaged flow quantities, statistical quantities, and density schlieren. Subsequently, spectral proper orthogonal decomposition analysis was utilized to effectively assess the coherent flow structures present in wavepackets. The results demonstrate that the inlet pressure significantly affects the flow characteristics. In particular, jet morphology variations manifest as an increased shear layer angle at the orifice throat and changes in the spacing of shock cells. The jet core and the wake dissipation region are significantly extended with the rise of the inlet pressure. For Pin = 3 bars, a distinct subsonic region with complex vortex structures is found behind the Mach disk. When Pin = 1 bar, the noise originates from the Kelvin–Helmholtz (K–H) instability and vortex shedding, with the K–H instability dominating at high frequencies. When Pin = 2 bars, the K–H instability is the dominant contributor to the noise. In contrast to the previous two conditions, the low-frequency double-layer wavepacket structure dominates when Pin = 3 bars, exhibiting positive and negative antisymmetric alternating motion.
- 1 February 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38 (2)
Delayed detached-eddy simulation in ducted orifices jet dynamics for low-pressure compressed air energy storage system
In this study, turbulent flow through an orifice was numerically modeled using delayed detached-eddy simulations to clarify transient flow dynamics and coherent flow structures. Three inlet pressures (Pin = 1 bar, 2 bars, and 3 bars) were selected for comparative analysis under a typical orifice porosity (orifice hole area to pipe area ratio) of β = 28.4%. The numerical results agreed well with the previous experimental measurements. The unsteady jet characteristics were first investigated regarding the time-averaged flow quantities, statistical quantities, and density schlieren. Subsequently, spectral proper orthogonal decomposition analysis was utilized to effectively assess the coherent flow structures present in wavepackets. The results demonstrate that the inlet pressure significantly affects the flow characteristics. In particular, jet morphology variations manifest as an increased shear layer angle at the orifice throat and changes in the spacing of shock cells. The jet core and the wake dissipation region are significantly extended with the rise of the inlet pressure. For Pin = 3 bars, a distinct subsonic region with complex vortex structures is found behind the Mach disk. When Pin = 1 bar, the noise originates from the Kelvin–Helmholtz (K–H) instability and vortex shedding, with the K–H instability dominating at high frequencies. When Pin = 2 bars, the K–H instability is the dominant contributor to the noise. In contrast to the previous two conditions, the low-frequency double-layer wavepacket structure dominates when Pin = 3 bars, exhibiting positive and negative antisymmetric alternating motion.
- 21 January 2026
- Published by Elsevier in Ocean Engineering
- Vol. 351
Unsteady flow behavior and fluid-induced noise characteristics of T-junction pipe with a closed main branch
- 21 January 2026
- Published by Elsevier in Ocean Engineering
- Vol. 351
Unsteady flow behavior and fluid-induced noise characteristics of T-junction pipe with a closed main branch
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (9)
Unsteady flow behaviors and coherent flow structures in closed branch T-junctions: Time-resolved refractive index matched particle image velocimetry measurements
This study experimentally investigates unsteady flow behavior and coherent structures in T-junctions with a closed branch, employing time-resolved refractive index matched particle image velocimetry (TR-RIM-PIV). The effects of three closed branch depths (L/D= 1, 2, and 4) on flow characteristics were analyzed for both closed straight and side branches at a Reynolds number of 10 000. High-resolution spatiotemporal flow features were captured using TR-RIM-PIV, while proper orthogonal decomposition revealed the energetic flow structures. Phase-dependent variations of these structures were examined to establish correlations between flow dynamics and noise generation. In the closed straight branch configuration (model 1), large-scale vortex structures emerged due to flow separation at both the leading edge and the opposite wall. The downstream shedding of the leading-edge vortex was driven by Kelvin–Helmholtz instability. For L/D= 1, the lateral jet intermittently impacted the rear wall, intensifying pressure fluctuations and converting hydrodynamic energy into acoustic energy. As the cavity depth increased, vortex–wall interactions weakened, resulting in one, two, and three recirculation zones for L/D= 1, 2, and 4, respectively, with decreasing intensity and size. In the closed side branch configuration (model 2), the turning jet adhered to the wall due to the Coanda effect. At the downstream corner, the jet split into two oppositely phased components, forming double-wavepacket structures that intermittently shed and propagated downstream. For L/D= 1, a single clockwise recirculation zone formed. For L/D= 2 and 4, deeper jet penetration produced additional counterclockwise and clockwise recirculation zones.
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (9)
Unsteady flow behaviors and coherent flow structures in closed branch T-junctions: Time-resolved refractive index matched particle image velocimetry measurements
This study experimentally investigates unsteady flow behavior and coherent structures in T-junctions with a closed branch, employing time-resolved refractive index matched particle image velocimetry (TR-RIM-PIV). The effects of three closed branch depths (L/D= 1, 2, and 4) on flow characteristics were analyzed for both closed straight and side branches at a Reynolds number of 10 000. High-resolution spatiotemporal flow features were captured using TR-RIM-PIV, while proper orthogonal decomposition revealed the energetic flow structures. Phase-dependent variations of these structures were examined to establish correlations between flow dynamics and noise generation. In the closed straight branch configuration (model 1), large-scale vortex structures emerged due to flow separation at both the leading edge and the opposite wall. The downstream shedding of the leading-edge vortex was driven by Kelvin–Helmholtz instability. For L/D= 1, the lateral jet intermittently impacted the rear wall, intensifying pressure fluctuations and converting hydrodynamic energy into acoustic energy. As the cavity depth increased, vortex–wall interactions weakened, resulting in one, two, and three recirculation zones for L/D= 1, 2, and 4, respectively, with decreasing intensity and size. In the closed side branch configuration (model 2), the turning jet adhered to the wall due to the Coanda effect. At the downstream corner, the jet split into two oppositely phased components, forming double-wavepacket structures that intermittently shed and propagated downstream. For L/D= 1, a single clockwise recirculation zone formed. For L/D= 2 and 4, deeper jet penetration produced additional counterclockwise and clockwise recirculation zones.
- 29 January 2025
- Published by World Scientific Pub Co Pte Ltd in International Journal of Modern Physics C
- Vol. 36 (11)
Numerical investigation on the differential injection angle effects of intrathecal anesthetic drug delivery
In lumbar anesthesia surgery, the accurate determination of key parameters such as injection angle and dosage is facilitated by the real-time observation of intrathecal drug distribution. In this study, a non-invasive and visual computational fluid dynamics method was used to simulate drug injections in the “cloverleaf” lumbar thoracic segment with a multiphase mixture model, clarifying the intrathecal unsteady flow behaviors and the drug transport mechanisms. The transport properties of three clinically significant injection angles — [ ], [ ] and [ ]— were compared, determining the fastest and slowest block onset angles based on drug concentration levels in the thoracic vertebral plane. The calculation results are aligned with clinical practice trends, and it is found that significantly different drug amounts are required for different injection angles, with higher doses necessitated for ultrasound-assisted anesthesia. During anesthesia, a trend of increasing and then decreasing local anesthetic concentrations in each vertebral plane was observed, with the presence of a minimum effective dose of local anesthetic at the target block site. Injecting at the “concave” part of the cloverleaf ([ ] and [ ]) results in greater drug pulsation, faster diffusion and more efficient mixing than injecting at the positive “convex” part of the cloverleaf cross-section ([ ]). At the L2 vertebral plane, the flow rate and mixing ratios during injection for angles of [ ], [ ] and [ ] were approximately 1:4:2. Compared with orthostatic [ ] injections, [ ] and [ ] oblique injections resulted in higher drug concentrations in the T12-10 vertebral plane during the pre-transport period, with [ ] injections being the fastest. However, over a longer period of time, [ ] injections resulted in the lowest drug concentrations in the target block plane (T10), necessitating a higher dose. A faster anesthetic response was produced by [ ] injections at the same dose. Valuable insights for the application of appropriate drug dosages in clinical anesthesia practice are offered by the calculations.
- 29 January 2025
- Published by World Scientific Pub Co Pte Ltd in International Journal of Modern Physics C
- Vol. 36 (11)
Numerical investigation on the differential injection angle effects of intrathecal anesthetic drug delivery
In lumbar anesthesia surgery, the accurate determination of key parameters such as injection angle and dosage is facilitated by the real-time observation of intrathecal drug distribution. In this study, a non-invasive and visual computational fluid dynamics method was used to simulate drug injections in the “cloverleaf” lumbar thoracic segment with a multiphase mixture model, clarifying the intrathecal unsteady flow behaviors and the drug transport mechanisms. The transport properties of three clinically significant injection angles — [ ], [ ] and [ ]— were compared, determining the fastest and slowest block onset angles based on drug concentration levels in the thoracic vertebral plane. The calculation results are aligned with clinical practice trends, and it is found that significantly different drug amounts are required for different injection angles, with higher doses necessitated for ultrasound-assisted anesthesia. During anesthesia, a trend of increasing and then decreasing local anesthetic concentrations in each vertebral plane was observed, with the presence of a minimum effective dose of local anesthetic at the target block site. Injecting at the “concave” part of the cloverleaf ([ ] and [ ]) results in greater drug pulsation, faster diffusion and more efficient mixing than injecting at the positive “convex” part of the cloverleaf cross-section ([ ]). At the L2 vertebral plane, the flow rate and mixing ratios during injection for angles of [ ], [ ] and [ ] were approximately 1:4:2. Compared with orthostatic [ ] injections, [ ] and [ ] oblique injections resulted in higher drug concentrations in the T12-10 vertebral plane during the pre-transport period, with [ ] injections being the fastest. However, over a longer period of time, [ ] injections resulted in the lowest drug concentrations in the target block plane (T10), necessitating a higher dose. A faster anesthetic response was produced by [ ] injections at the same dose. Valuable insights for the application of appropriate drug dosages in clinical anesthesia practice are offered by the calculations.
- 1 December 2024
- Published by AIP Publishing in Physics of Fluids
- Vol. 36 (12)
Flow behaviors and throttle effects of high-pressure air flow through orifices
In this study, the flow behaviors and throttle effects of high-pressure air flow through orifices were experimentally investigated in a pipe system and numerically simulated using Reynolds-averaged Navier–Stokes equations with the shear stress transport (SST) k−ω turbulence model. Three orifice configurations with porosities (orifice hole area to pipe area ratio) of β = 28.4%, 7.1%, and 1.8% were comparatively evaluated under four different inflow pressures, i.e., pin= 1, 2, 3, and 4 bar. The numerical results agreed well with the corresponding experimental measurements. A primary recirculation with a smaller secondary recirculation region was formed immediately behind the trailing face. In the presence of supersonic flow, the streamwise velocity exhibited periodic oscillations, increasing at the expansion wave and decreasing at the shock wave region. The throttle effects were highly sensitive to the orifice porosities and the inflow conditions. In the β = 28.4% configuration, shock waves and expansion waves were evident under different inflow pressures and alternately dominated the downstream flow, leading to periodic streamwise temperature variations, with the minimum temperature reaching −160 °C at pin = 4 bar. For the orifice with β = 7.1%, the expansion and shock waves appeared at pin≥ 2 bar, and continuously expanded with intensified flow fluctuations at pin = 3 and 4 bar, with the minimum temperature reaching −120 °C. For the orifice with β = 1.8%, the impact of shock waves on the temperature field was barely observable under pin = 1 and 2 bar due to the small porosity. When pin≥ 3 bar, similar temperature patterns with larger-porosity orifices were observed downstream of the orifice trailing face but were smaller in size.
- 1 December 2024
- Published by AIP Publishing in Physics of Fluids
- Vol. 36 (12)
Flow behaviors and throttle effects of high-pressure air flow through orifices
In this study, the flow behaviors and throttle effects of high-pressure air flow through orifices were experimentally investigated in a pipe system and numerically simulated using Reynolds-averaged Navier–Stokes equations with the shear stress transport (SST) k−ω turbulence model. Three orifice configurations with porosities (orifice hole area to pipe area ratio) of β = 28.4%, 7.1%, and 1.8% were comparatively evaluated under four different inflow pressures, i.e., pin= 1, 2, 3, and 4 bar. The numerical results agreed well with the corresponding experimental measurements. A primary recirculation with a smaller secondary recirculation region was formed immediately behind the trailing face. In the presence of supersonic flow, the streamwise velocity exhibited periodic oscillations, increasing at the expansion wave and decreasing at the shock wave region. The throttle effects were highly sensitive to the orifice porosities and the inflow conditions. In the β = 28.4% configuration, shock waves and expansion waves were evident under different inflow pressures and alternately dominated the downstream flow, leading to periodic streamwise temperature variations, with the minimum temperature reaching −160 °C at pin = 4 bar. For the orifice with β = 7.1%, the expansion and shock waves appeared at pin≥ 2 bar, and continuously expanded with intensified flow fluctuations at pin = 3 and 4 bar, with the minimum temperature reaching −120 °C. For the orifice with β = 1.8%, the impact of shock waves on the temperature field was barely observable under pin = 1 and 2 bar due to the small porosity. When pin≥ 3 bar, similar temperature patterns with larger-porosity orifices were observed downstream of the orifice trailing face but were smaller in size.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.2
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.2
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.1
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.1
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.3
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.
- 18 January 2023
- Published by AIP Publishing
- Vol. 35 (1) , 015142
10.1063/5.0136760.3
Turbulent flow through tandem orifices in a circular duct is numerically modeled through dynamic delayed detached-eddy simulations to clarify the unsteady flow behaviors and noise generation mechanism. The characteristics of four configurations with different separation distances L/D = 0 (single orifice), 1, 2, and 4 are compared at a Reynolds number of 10 000. The acoustic sources and their noise-propagation behaviors are analyzed using Lighthill's acoustic analogy. The coherent flow structures (wavepackets) are determined through spectral proper orthogonal decomposition to clarify the resulting flow noise mechanism. The dominant noise sources are acoustic dipoles that are alternately energetic on the orifice's leading and trailing faces subjected to intermittent interaction with the unsteady flow. The total sound pressure level (SPL) for a single orifice is alternately dominated by the shedding and flapping behaviors of the large-scale vortical structures in the low-frequency range and Kelvin–Helmholtz (K–H) type wavepackets at high frequencies. For the tandem orifice configuration with L/D = 1, the total SPL is dominated by the contribution of the trailing face, attributable to the interactions between the K–H-type double-wavepacket structures. Both the upstream and downstream wavepackets start to split and generate four-wavepacket structures in the high-frequency range. In the cases of L/D = 2 and 4, the total SPL is dominated by the dipole sources at the downstream leading face that subjected to the intermittent interaction with the upstream separated shear layers. The dominant flow structures are independent wavepackets through each orifice at low frequencies, while at high frequencies the four-wavepacket structures break into two independent double-wavepacket structures close to each orifice.