Joseph Yen - Academia.edu (original) (raw)

Papers by Joseph Yen

45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2009

25th AIAA/CEAS Aeroacoustics Conference

: The purpose of the testing reported herein was to obtain data describing fluid and structural b... more : The purpose of the testing reported herein was to obtain data describing fluid and structural behavior (interaction of fluid on structure and of structure on fluid) in a controlled experiment in a systematic manner. These data are to be used for (a) validation of fluid/structure interaction (FSI) modeling codes and (b) to demonstrate a process for archiving and accessing the data. The approach, which uses two types of models (wing and tethered mass), was to obtain ambient upstream, wall, and downstream static and dynamic fluid pressures for the models in various dynamic conditions.

26th AIAA Applied Aerodynamics Conference, 2008

A large number of papers have been published on supersonic nozzle design since Prandtl and Busema... more A large number of papers have been published on supersonic nozzle design since Prandtl and Buseman first applied the Method of Characteristics in a paper published in 1929. This study explores a method by Evvard and Marcus published in 1952 for 2D supersonic wind tunnel nozzle designs. Based on the Method of Characteristics, this nozzle design method achieves continuous wall curvature avoiding disturbances to the supersonic flow field and providing the curvature continuity needed for variable Mach number flexible-walled nozzles. In the current study, the method is found to be capable of creating perfectly uniform test section air flow for a fixed-block nozzle design for supersonic Mach numbers ≤ 5.5. Moreover, simultaneously achieving perfect flow uniformity, smooth curvatures, and single inflection point, the nozzle contours for the range of Mach numbers can span a common length which is required in a flexible-walled nozzle. In addition, the current study also develops a visualization capability to inspect the characteristics network and resultant nozzle flow field. It is discovered that as the accelerating Mach numbers turn constant and uniform in the rhombus region, the accompanying compression (right-running characteristic) waves become perfectly linear and their reflections remain so downstream. No wave cancellation is implemented or observed, which differs from the description in other design methods.

17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference), 2011

12th AIAA/CEAS Aeroacoustics Conference (27th AIAA Aeroacoustics Conference), 2006

In this paper, 3D CAA solutions are obtained using an extension of Courant number insensitive con... more In this paper, 3D CAA solutions are obtained using an extension of Courant number insensitive conservation element and solution element (CESE) method first introduced by SC. Chang. A Courant number insensitive scheme (CNIS) retains the solution quality even in a flow simulation with a large disparity in the grid Courant numbers (CFL). However, implementing CNIS incurs extra computational cost due to the grid CFL calculation. For a simulation applying unstructured meshes, this cost becomes prohibitive if the evaluation is based on the formal CFL definition. Recently, a 2D CNIS with simplified grid CFL calculation was shown to yield accurate solutions for a range of CAA and CFD problems. The current study extends the simplified CNIS to a 3D version. It is shown that the 3D version is capable of accurately preserving the wave solution quality including wave form and amplitude for a CAA catagory-3 benchmark problem. In addition, the new scheme also yields improved solution for a 3D oblique shock problem.

45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2009

This paper is a parallel effort to another (AIAA 2009-5011) by the authors to study the heating c... more This paper is a parallel effort to another (AIAA 2009-5011) by the authors to study the heating condition in the A3 subscale rocket test facility diffuser during a hot fire test. The aim of both papers is to reveal the maximum heating values which can be used for the design of the full-scale A3 diffuser structures. This paper reports the use of Cook-Feldman method for calculating heat flux from experimental surface temperature measurements taken on the A3 subscale diffuser during firing of the subscale rocket engine. The Cook-Feldman algorithm, based on surface temperature history, provides an efficient means of heat flux calculation for batch processing the large temperature data files from the comprehensive scale diffuser test program. The study reveals the transient character of heating process at various representative locations along the subscale diffuser from which maximum local heat flux values are identified. Among the instrumented locations, the elbow area of the diffuser ductwork emerges as the most critical heating location. The study also reveals the spatial variation of the heat transfer across the high-speed plume contact with the capture duct of the diffuser. The transient heating estimate accompanying the supersonic flow recompression is shown.

43rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2007

21st AIAA Computational Fluid Dynamics Conference, 2013

SAE International Journal of Passenger Cars - Mechanical Systems, 2013

U.S. Air Force T&E Days 2010, 2010

This study demonstrates the analysis capability of an emerging Computational Aeroacoustics (CAA) ... more This study demonstrates the analysis capability of an emerging Computational Aeroacoustics (CAA) method by applying the method to the Weapons Internal Carriage and Separation (WICS) experimental database produced at Arnold Engineering Development Center (AEDC). The applied analysis tool is based on the numerical method of space-time Conservation Element and Solution Element (CESE), which is a nontraditional, highresolution, unstructured-mesh CAA method satisfied through second-order accuracy in space and time compared to other higher-order methods. The CESE method is not based on an upwind scheme and does not use a Riemann solver in the solution formulation. The resolution of the large-scale eddies of the cavity shear layer is achieved using a turbulence modeling method of Implicit Large Eddy Simulation (ILES) in which the numerical dissipations, including truncation errors and solution gradient-adaptive damping of the CESE method, takes the place of an explicit sub-grid scale (SGS) model, typically applied in an LES. The demonstration study includes the simplified 2D and physical 3D geometry of an open cavity of 4.5 length-to-depth ratio ( L/D ) that is subject to supersonic freestream conditions of Mach numbers from 1.2 to 5.04. The study shows that the 3D CESE solution reveals a frequency lockup condition inside the cavity where the predicted tones and amplitudes are very agreeable to the experiment. Good agreement is also seen in the broadband frequency spectrum even in the thousand Hertz frequencies which is rather striking from a 2 nd -order accurate solution method. The result also suggests good modeling capability of the ILES method for cavity problems. Due to the simplified resolution of the turbulent flow, the 2D study shows different prediction quality which is better at lower Mach numbers and progressively more deviated at higher Mach numbers. Nonetheless, the lower-cost 2D method can still provide a first-order estimate of the potential tones and amplitudes whereas the latter information is absent from an analytical method, such as the Rossiter’s model.

18th AIAA/CEAS Aeroacoustics Conference (33rd AIAA Aeroacoustics Conference), 2012

This paper reports a numerical study of low-frequency acoustics on a modern open-jet wind tunnel ... more This paper reports a numerical study of low-frequency acoustics on a modern open-jet wind tunnel design. The study simulates the acoustic response inside the wind tunnel which is subjected to a simulated edge-tone emission without a moving flow. The effects of a Helmholtz resonator installed at different locations along the wind tunnel circuit are investigated with the aim to attenuate the resonance effect of the edge-tone noise. The simulation assumes that the frequency of the edge tone is equal to the natural frequency of the resonator and the corresponding wavelength is the same as the total longitudinal length (~ 72.4 m) along the wind tunnel circuit centerline. Due to the long wavelength, the study excludes the relatively small length-scale components, including corner turning vanes, main fan, and main fan hub, from the wind tunnel geometry. The computational aeroacoustics study is performed using a solver JUSTUS which is based on a time-accurate local-time-stepping space-time conservation element and solution element method in which the full (instead of linearized) Euler equations are solved. All solutions are obtained using unstructured tetrahedral meshes with local mesh refinements, which are required for accurate geometry representation. The design of the unstructured meshes results in a very large disparity between the minimum and maximum local time step sizes whose ratio is more than three orders in magnitude. The study starts with a demonstration problem of a standalone resonator on a straight duct, and the prediction is compared to a lumped-parameter theory. Resulting in agreeable solution to the theory, the similar mesh design for the standalone resonator is then applied to the application study for the wind tunnel. The study shows that, without a resonator, an acoustic resonance occurs in the wind tunnel circuit where the response pressure magnitudes are higher than the source strength. Two acoustic nodes (of locally lowest pressure magnitudes) are predicted in the baseline wind tunnel, and this result is shown to agree to a 1D theory which is developed in the study. The pattern and response amplitude of the resonance are modified with a resonator. It is found that, among the resonator cases, the resonator atop the test section creates the broadest area of constant-level attenuation. Different resonator locations result in different levels of attenuation along the wind tunnel circuit. It is surprising to find that the case of a resonator on corner 3 increases, albeit incrementally, the response magnitudes in comparison to the baseline condition. The results suggest that a resonator may be strategically located to produce one acoustic node at or near the nozzle exit plane where the initial jet shear-layer is most susceptible to acoustic excitations. It is expected that the lowest excitation would minimize the tendency of the jet instability to become coherently amplified and develop into an edge-tone feedback intensification. The resonator location may effectively attenuate the edge-tone feedback phenomenon in the test section and thus diminish the resulting strength of resonance with the circuit. As an initial effort, the current acoustic study neglects the effects of the moving mean flow and assumes the low-frequency noise to emit at fixed locations and be statistically stationary. A more accurate description of the acoustics, influenced by convection effect and flow-sound coupling, is to allow the physical aeroacoustics to take place in the wind tunnel with the moving mean flow. This will be a continuous effort of the computational study.

42nd AIAA Aerospace Sciences Meeting and Exhibit, 2004

SAE Technical Paper Series, 2003

Computational Fluid Dynamics (CFD) was applied to investigate blockage effects (or velocity corre... more Computational Fluid Dynamics (CFD) was applied to investigate blockage effects (or velocity correction) in a climatic wind tunnel (CWT) test environment. Different blockage effects in the CWT were modeled using four simplified vehicles that approximated a sedan, an SUV, a pickup truck, and a minivan. Blockage dependence on nozzle size and spacing between the nozzle exit plane (NEP) and the vehicle were also investigated. The study quantified the blockage effect using different correction methods based on vehicle frontal velocity profiles and upper surface pressure traces. The blockage-free solution was also simulated for each vehicle in an 'open road' or free air condition. The CFD study revealed that all the test cases resulted in blockage correction factors, defined by V a c t u a l /V s i m u l a t e d greater than 1.0. This is a condition in which the uncorrected wind tunnel velocity was higher than the 'open road' condition. It was determined that vehicles with higher C D A F values induced higher blockage correction factors. The blockage correction factor increased exponentially as the vehicle was positioned closer to the NEP. Small differences in the blockage factor between the frontal velocity profile method and the upper vehicle surface pressure trace method were found. These results led to two unified CWT blockage correlations based on the vehicle frontal velocity profiles and the upper surface pressure traces. Each correlation is valid across a range of vehicle frontal areas, drag coefficients, and separation distances between the nozzle exit plane and the vehicle.

AIAA Journal, 1998

A newly developed method called parabolized stability equations (PSE) is used to analyze the stre... more A newly developed method called parabolized stability equations (PSE) is used to analyze the streamwise evolution of instability waves in flows that are highly unidirectional, such as boundary layers and free shear layers. That the PSE analysis is able to fully account for the nature of diverging mean flow and to resolve the evolving instabilities without numerical difficulties as seen in linear stability theory has made PSE an alternative tool in dealing with the jet instabilities at relatively low computational cost compared with schemes based on direct numerical simulation or large eddy simulation

8th AIAA/CEAS Aeroacoustics Conference & Exhibit, 2002

11 th AIAA/CEAS Aeroacoustics Conference(26 th …, 2005

Computational Aeroacoustics Using a Simplified Courant Number Insensitive CE/SE Method. Joseph C ... more Computational Aeroacoustics Using a Simplified Courant Number Insensitive CE/SE Method. Joseph C Yen, Donald A Wagner 11 th AIAA/CEAS Aeroacoustics Conference(26 th Aeroacoustics Conference); Monterey, CA, 1 ...

SAE Technical Paper Series, 2004

45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2009

25th AIAA/CEAS Aeroacoustics Conference

: The purpose of the testing reported herein was to obtain data describing fluid and structural b... more : The purpose of the testing reported herein was to obtain data describing fluid and structural behavior (interaction of fluid on structure and of structure on fluid) in a controlled experiment in a systematic manner. These data are to be used for (a) validation of fluid/structure interaction (FSI) modeling codes and (b) to demonstrate a process for archiving and accessing the data. The approach, which uses two types of models (wing and tethered mass), was to obtain ambient upstream, wall, and downstream static and dynamic fluid pressures for the models in various dynamic conditions.

26th AIAA Applied Aerodynamics Conference, 2008

A large number of papers have been published on supersonic nozzle design since Prandtl and Busema... more A large number of papers have been published on supersonic nozzle design since Prandtl and Buseman first applied the Method of Characteristics in a paper published in 1929. This study explores a method by Evvard and Marcus published in 1952 for 2D supersonic wind tunnel nozzle designs. Based on the Method of Characteristics, this nozzle design method achieves continuous wall curvature avoiding disturbances to the supersonic flow field and providing the curvature continuity needed for variable Mach number flexible-walled nozzles. In the current study, the method is found to be capable of creating perfectly uniform test section air flow for a fixed-block nozzle design for supersonic Mach numbers ≤ 5.5. Moreover, simultaneously achieving perfect flow uniformity, smooth curvatures, and single inflection point, the nozzle contours for the range of Mach numbers can span a common length which is required in a flexible-walled nozzle. In addition, the current study also develops a visualization capability to inspect the characteristics network and resultant nozzle flow field. It is discovered that as the accelerating Mach numbers turn constant and uniform in the rhombus region, the accompanying compression (right-running characteristic) waves become perfectly linear and their reflections remain so downstream. No wave cancellation is implemented or observed, which differs from the description in other design methods.

17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference), 2011

12th AIAA/CEAS Aeroacoustics Conference (27th AIAA Aeroacoustics Conference), 2006

In this paper, 3D CAA solutions are obtained using an extension of Courant number insensitive con... more In this paper, 3D CAA solutions are obtained using an extension of Courant number insensitive conservation element and solution element (CESE) method first introduced by SC. Chang. A Courant number insensitive scheme (CNIS) retains the solution quality even in a flow simulation with a large disparity in the grid Courant numbers (CFL). However, implementing CNIS incurs extra computational cost due to the grid CFL calculation. For a simulation applying unstructured meshes, this cost becomes prohibitive if the evaluation is based on the formal CFL definition. Recently, a 2D CNIS with simplified grid CFL calculation was shown to yield accurate solutions for a range of CAA and CFD problems. The current study extends the simplified CNIS to a 3D version. It is shown that the 3D version is capable of accurately preserving the wave solution quality including wave form and amplitude for a CAA catagory-3 benchmark problem. In addition, the new scheme also yields improved solution for a 3D oblique shock problem.

45th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2009

This paper is a parallel effort to another (AIAA 2009-5011) by the authors to study the heating c... more This paper is a parallel effort to another (AIAA 2009-5011) by the authors to study the heating condition in the A3 subscale rocket test facility diffuser during a hot fire test. The aim of both papers is to reveal the maximum heating values which can be used for the design of the full-scale A3 diffuser structures. This paper reports the use of Cook-Feldman method for calculating heat flux from experimental surface temperature measurements taken on the A3 subscale diffuser during firing of the subscale rocket engine. The Cook-Feldman algorithm, based on surface temperature history, provides an efficient means of heat flux calculation for batch processing the large temperature data files from the comprehensive scale diffuser test program. The study reveals the transient character of heating process at various representative locations along the subscale diffuser from which maximum local heat flux values are identified. Among the instrumented locations, the elbow area of the diffuser ductwork emerges as the most critical heating location. The study also reveals the spatial variation of the heat transfer across the high-speed plume contact with the capture duct of the diffuser. The transient heating estimate accompanying the supersonic flow recompression is shown.

43rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2007

21st AIAA Computational Fluid Dynamics Conference, 2013

SAE International Journal of Passenger Cars - Mechanical Systems, 2013

U.S. Air Force T&E Days 2010, 2010

This study demonstrates the analysis capability of an emerging Computational Aeroacoustics (CAA) ... more This study demonstrates the analysis capability of an emerging Computational Aeroacoustics (CAA) method by applying the method to the Weapons Internal Carriage and Separation (WICS) experimental database produced at Arnold Engineering Development Center (AEDC). The applied analysis tool is based on the numerical method of space-time Conservation Element and Solution Element (CESE), which is a nontraditional, highresolution, unstructured-mesh CAA method satisfied through second-order accuracy in space and time compared to other higher-order methods. The CESE method is not based on an upwind scheme and does not use a Riemann solver in the solution formulation. The resolution of the large-scale eddies of the cavity shear layer is achieved using a turbulence modeling method of Implicit Large Eddy Simulation (ILES) in which the numerical dissipations, including truncation errors and solution gradient-adaptive damping of the CESE method, takes the place of an explicit sub-grid scale (SGS) model, typically applied in an LES. The demonstration study includes the simplified 2D and physical 3D geometry of an open cavity of 4.5 length-to-depth ratio ( L/D ) that is subject to supersonic freestream conditions of Mach numbers from 1.2 to 5.04. The study shows that the 3D CESE solution reveals a frequency lockup condition inside the cavity where the predicted tones and amplitudes are very agreeable to the experiment. Good agreement is also seen in the broadband frequency spectrum even in the thousand Hertz frequencies which is rather striking from a 2 nd -order accurate solution method. The result also suggests good modeling capability of the ILES method for cavity problems. Due to the simplified resolution of the turbulent flow, the 2D study shows different prediction quality which is better at lower Mach numbers and progressively more deviated at higher Mach numbers. Nonetheless, the lower-cost 2D method can still provide a first-order estimate of the potential tones and amplitudes whereas the latter information is absent from an analytical method, such as the Rossiter’s model.

18th AIAA/CEAS Aeroacoustics Conference (33rd AIAA Aeroacoustics Conference), 2012

This paper reports a numerical study of low-frequency acoustics on a modern open-jet wind tunnel ... more This paper reports a numerical study of low-frequency acoustics on a modern open-jet wind tunnel design. The study simulates the acoustic response inside the wind tunnel which is subjected to a simulated edge-tone emission without a moving flow. The effects of a Helmholtz resonator installed at different locations along the wind tunnel circuit are investigated with the aim to attenuate the resonance effect of the edge-tone noise. The simulation assumes that the frequency of the edge tone is equal to the natural frequency of the resonator and the corresponding wavelength is the same as the total longitudinal length (~ 72.4 m) along the wind tunnel circuit centerline. Due to the long wavelength, the study excludes the relatively small length-scale components, including corner turning vanes, main fan, and main fan hub, from the wind tunnel geometry. The computational aeroacoustics study is performed using a solver JUSTUS which is based on a time-accurate local-time-stepping space-time conservation element and solution element method in which the full (instead of linearized) Euler equations are solved. All solutions are obtained using unstructured tetrahedral meshes with local mesh refinements, which are required for accurate geometry representation. The design of the unstructured meshes results in a very large disparity between the minimum and maximum local time step sizes whose ratio is more than three orders in magnitude. The study starts with a demonstration problem of a standalone resonator on a straight duct, and the prediction is compared to a lumped-parameter theory. Resulting in agreeable solution to the theory, the similar mesh design for the standalone resonator is then applied to the application study for the wind tunnel. The study shows that, without a resonator, an acoustic resonance occurs in the wind tunnel circuit where the response pressure magnitudes are higher than the source strength. Two acoustic nodes (of locally lowest pressure magnitudes) are predicted in the baseline wind tunnel, and this result is shown to agree to a 1D theory which is developed in the study. The pattern and response amplitude of the resonance are modified with a resonator. It is found that, among the resonator cases, the resonator atop the test section creates the broadest area of constant-level attenuation. Different resonator locations result in different levels of attenuation along the wind tunnel circuit. It is surprising to find that the case of a resonator on corner 3 increases, albeit incrementally, the response magnitudes in comparison to the baseline condition. The results suggest that a resonator may be strategically located to produce one acoustic node at or near the nozzle exit plane where the initial jet shear-layer is most susceptible to acoustic excitations. It is expected that the lowest excitation would minimize the tendency of the jet instability to become coherently amplified and develop into an edge-tone feedback intensification. The resonator location may effectively attenuate the edge-tone feedback phenomenon in the test section and thus diminish the resulting strength of resonance with the circuit. As an initial effort, the current acoustic study neglects the effects of the moving mean flow and assumes the low-frequency noise to emit at fixed locations and be statistically stationary. A more accurate description of the acoustics, influenced by convection effect and flow-sound coupling, is to allow the physical aeroacoustics to take place in the wind tunnel with the moving mean flow. This will be a continuous effort of the computational study.

42nd AIAA Aerospace Sciences Meeting and Exhibit, 2004

SAE Technical Paper Series, 2003

Computational Fluid Dynamics (CFD) was applied to investigate blockage effects (or velocity corre... more Computational Fluid Dynamics (CFD) was applied to investigate blockage effects (or velocity correction) in a climatic wind tunnel (CWT) test environment. Different blockage effects in the CWT were modeled using four simplified vehicles that approximated a sedan, an SUV, a pickup truck, and a minivan. Blockage dependence on nozzle size and spacing between the nozzle exit plane (NEP) and the vehicle were also investigated. The study quantified the blockage effect using different correction methods based on vehicle frontal velocity profiles and upper surface pressure traces. The blockage-free solution was also simulated for each vehicle in an 'open road' or free air condition. The CFD study revealed that all the test cases resulted in blockage correction factors, defined by V a c t u a l /V s i m u l a t e d greater than 1.0. This is a condition in which the uncorrected wind tunnel velocity was higher than the 'open road' condition. It was determined that vehicles with higher C D A F values induced higher blockage correction factors. The blockage correction factor increased exponentially as the vehicle was positioned closer to the NEP. Small differences in the blockage factor between the frontal velocity profile method and the upper vehicle surface pressure trace method were found. These results led to two unified CWT blockage correlations based on the vehicle frontal velocity profiles and the upper surface pressure traces. Each correlation is valid across a range of vehicle frontal areas, drag coefficients, and separation distances between the nozzle exit plane and the vehicle.

AIAA Journal, 1998

A newly developed method called parabolized stability equations (PSE) is used to analyze the stre... more A newly developed method called parabolized stability equations (PSE) is used to analyze the streamwise evolution of instability waves in flows that are highly unidirectional, such as boundary layers and free shear layers. That the PSE analysis is able to fully account for the nature of diverging mean flow and to resolve the evolving instabilities without numerical difficulties as seen in linear stability theory has made PSE an alternative tool in dealing with the jet instabilities at relatively low computational cost compared with schemes based on direct numerical simulation or large eddy simulation

8th AIAA/CEAS Aeroacoustics Conference & Exhibit, 2002

11 th AIAA/CEAS Aeroacoustics Conference(26 th …, 2005

Computational Aeroacoustics Using a Simplified Courant Number Insensitive CE/SE Method. Joseph C ... more Computational Aeroacoustics Using a Simplified Courant Number Insensitive CE/SE Method. Joseph C Yen, Donald A Wagner 11 th AIAA/CEAS Aeroacoustics Conference(26 th Aeroacoustics Conference); Monterey, CA, 1 ...

SAE Technical Paper Series, 2004