National Aeronautics and Space Administration

www.nasa.gov
Computational Analysis of a Chevron Nozzle Uniquely
Tailored for Propulsion Airframe Aeroacoustics
12
th
AIAA/CEAS Aeroacoustics Conference
Cambridge, MA
May 8-10, 2006
Steven J. Massey
Eagle Aeronautics, Inc.
Alaa A. Elmiligui
Analytical Services & Materials, Inc.
Craig A. Hunter, Russell H. Thomas, S. Paul Pao
NASA Langley Research Center
and
Vinod G. Mengle
Boeing Company
May 8, 2006 NASA Langley Research Center 2
Outline
Motivation
Objectives
Numerical Tools
Review of Generic Jet-Pylon Effect
Axi, bb, RR, RT Nozzle Configurations
Analysis Procedure
Results Chain from Noise to Geometry
Summary
Concluding Remarks
May 8, 2006 NASA Langley Research Center 3
General PAA Related Effects and Features
On Typical Conventional Aircraft
Nacelle-airframe integration
e.g. chines, flow distortion,
relative angles
Jet-pylon
interaction of the
PAA T-fan nozzle
Jet-flap
impingement
Jet-flap trailing
edge interaction
Jet influence on
airframe sources:
side edges
Jet interaction with
horizontal stabilizers
Jet and fan noise
scattering from
fuselage, wing, flap
surfaces
Pylon-slat cutout
QTD2 partnership of
Boeing, GE, Goodrich,
NASA, and ANA
May 8, 2006 NASA Langley Research Center 4
Objectives
To build a predictive capability to link geometry
to noise for complex configurations
To identify the flow and noise source
mechanisms of the PAA T-Fan (quieter at take
off than the reference chevron nozzle)
May 8, 2006 NASA Langley Research Center 5
Numerical Tools
PAB3D
3D RANS upwind code
Multi-block structured with general patching
Parallel using MPI
Mesh sequencing
Two-equation k-! turbulence models
Several algebraic Reynolds stress models
Jet3D
Lighthills Acoustic Analogy in 3D
Models the jet flow with a fictitious volume distribution
of quadrupole sources radiating into a uniform ambient
medium
Uses RANS CFD as input
Now implemented for structured and unstructured
grids (ref AIAA 2006-2597)
May 8, 2006 NASA Langley Research Center 6
Sample Grid Plane
31 Million Cells for 180
o
PAB3D solution: 33
hours on 44 Columbia
CPUs (Itanium 2)
Jet3D solution, 10
minutes on Mac
May 8, 2006 NASA Langley Research Center 7
Model Scale LSAF PAA Nozzles Analyzed
Four Nozzles Chosen for
Analysis:
Axisymmetric Nozzle
(not an experimental
nozzle)
bb conventional nozzles
RR state-of-the-art
azimuthally uniform
chevrons on core and
fan
RT PAA T-fan
azimuthally varying
chevrons on fan and
uniform chevrons on
core
For more details see
Mengle et al. AIAA 06-
2467
May 8, 2006 NASA Langley Research Center 8
Generic Pylon Effect Understanding - AIAA 05-3083
Core Flow Induced Off of Jet Axis by
Coanda Effect
Pairs of Large Scale Vortices Created
TKE and Noise Sources Move
Upstream
Depending on Design Details can
Result in Noise Reduction or Increase
with Pylon
Refs: AIAA 01-2183, 01-2185, 03-3169, 03-
3212, 04-2827, 05-3083
May 8, 2006 NASA Langley Research Center 9
Analysis Procedure
Start with established facts and work from
derived to fundamental quantities to form
connections to geometry
Measured noise data (LSAF)
SPL predictions (Jet3D)
OASPL noise source histogram (Jet3D)
Mass averaged, non-dimensional turbulence intensity
(PAB3D)
OASPL noise source maps (Jet3D)
Turbulence kinetic energy (PAB3D)
Axial vorticity
Cross flow streamlines
Vertical velocity
Total temperature
Total temperature centroid
Geometry
May 8, 2006 NASA Langley Research Center 10
Jet3D SPL Predictions with LSAF
*
* Axi case not
thrust matched
to others
Observer located on a 68.1D radius from the fan nozzle exit at an
inlet angle of 88.5 deg. and an azimuthal angle of 180 deg.
LSAF data from Mengle et al. AIAA 20062467
Tunnel
noise
bb predicted within 1 dB for
whole range
RR over predicted by 1 dB for
frequencies < 10 kHz, under
predicted by up to 2 dB for
high frequencies
RT predicted within 1 dB for
whole range, under predicted
high frequencies
Trends predicted
correctly increasing
confidence of flow
and noise source
linkage
May 8, 2006 NASA Langley Research Center 11
Noise Prediction CFD Link
Noise and TKE sources relative to Axi are consistent with previous
pylon understanding of mixing
Mass-Avg TKE qualitatively matches noise source histogram
bb, RR, RT intersect near x/D = 10
Axi crosses bb, RR at x/D = 12
Axi crosses RT at x/D = 12.75
Jet3D OASPL Histogram PAB3D: Mass-Avg TKE
May 8, 2006 NASA Langley Research Center 12
LAA CFD Correspondence
Axi bb RR RT
Peak noise
sources correspond
with peak TKE
Local noise
increased by
chevron length
Cross flow stream
lines show shear
layer vorticity
orientation
May 8, 2006 NASA Langley Research Center 13
Beginning Fan/Core Shear Merger
Noise and TKE peak
as layers merge
RR levels slightly
lower than bb
RT merger delayed,
much lower levels
Axi noise
asymmetry due to
LAA observer
location. TKE is
symmetric
Axial velocity 20
times stronger than
cross flow, thus
strongest vortex
would take about
60D for one
revolution
Axi bb RR RT
May 8, 2006 NASA Langley Research Center 14
Peak Noise From Shear Merger
bb, RR peak shown;
RT peaks 0.5D later,
one contour lower
than bb and RR
Unmerged Axi with
lower noise and TKE,
but will persist more
downstream
Axi bb RR RT
May 8, 2006 NASA Langley Research Center 15
Chevrons Add Vorticity
Axi cross flow is symmetric, so axial vorticity = zero
bb shows boundary layer vorticity shifted off axis by pylon
RT longer chevrons show increased vorticity over RR and
shorter chevrons on bottom show decreases
Plug
Core Cowl
P
y
l
o
n
May 8, 2006 NASA Langley Research Center 16
Pylon, Plug, Chevron Interaction
RT fan vortices more
defined on top, less
on bottom due to
chevron length
Vertical velocity
component shows
effect of pylon on
cross flow:
Axi shows Coanda
effect on plug
Pylon cases have
expanded downward
flow region to get
around pylon to fill
in plug
Less downward
movement in fan
flow for RT
May 8, 2006 NASA Langley Research Center 17
Consolidation and Entrainment
Core and fan shear
layer vorticity
consolidates to form
vortex pair
RR vortex pair
slightly stronger
than bb
RT vortex pair
significantly weaker
than bb and RR
May 8, 2006 NASA Langley Research Center 18
T-Fan Reduces Overall Mixing
RT local mixing
proportional to
chevron length
RT decreases net
mixing, extends core
by ~ 1/2 D
RR negligible mixing
over bb
May 8, 2006 NASA Langley Research Center 19
Overall Jet Trajectory
bb and RR equivalent symmetric chevron does not
interact with pylon effect
RT showing less downward movement favorable
interaction of asymmetric chevron with pylon effect
Total Temperature Centroid
May 8, 2006 NASA Langley Research Center 20
Summary
Overall mixing does not vary much between bb, RR
and RT and is not indicative of noise in this study
The T-Fan effect:
Varies the strength azimuthally of the localized
chevron vorticity
Reduces the downstream large scale vorticies
introduced by the pylon
Delays the merger of the fan and core shear layers
Reduces peak noise and shifts it downstream
There is the possibility of a more favorable design
for shear layer merger, which can now be found
computationally
May 8, 2006 NASA Langley Research Center 21
Concluding Remarks
A predictive capability linking geometry to noise
has been demonstrated
The T-Fan benefits from a favorable interaction
between asymmetric chevrons and the pylon effect
May 8, 2006 NASA Langley Research Center 22
Discussion, Extra Slides
May 8, 2006 NASA Langley Research Center 23
Axisymmetric Nozzle
Surfaces colored
by temperature
May 8, 2006 NASA Langley Research Center 24
Baseline Nozzle (bb)
Fan boundary
streamline
Near surface streamlines
and temperature
May 8, 2006 NASA Langley Research Center 25
Reference Chevrons (RR)
Slight upward
movement
Near surface streamlines
and temperature
May 8, 2006 NASA Langley Research Center 26
PAA T-Fan Nozzle (RT)
Near surface streamlines
and temperature
Further upward
movement
May 8, 2006 NASA Langley Research Center 27
Motivation
Propulsion Airframe Aeroacoustics (PAA)
Definition: Aeroacoustic effects associated with the
integration of the propulsion and airframe systems.
Includes:
Integration effects on inlet and exhaust systems
Flow interaction and acoustic propagation effects
Configurations from conventional to revolutionary
PAA goal is to reduce interaction effects directly or
use integration to reduce net radiated noise.
May 8, 2006 NASA Langley Research Center 28
PAA on QTD2: Concept to Flight in Two Years
Exploration of Possible PAA Concepts with
QTD2 Partners (5/03 4/04)
Extensive PAA CFD/Prediction Work (10/03
8/05)
(AIAA 05-3083, 06-2436)
PAA Experiment at Boeing LSAF
9/04
PAA Effects and Noise Reduction
Technologies Studied
AIAA 06-2467, 06-2434, 06-2435
PAA on QTD2 8/05
PAA T-Fan Chevron
Nozzle
PAA Effects
Instrumentation
AIAA 06-2438, 06-2439
May 8, 2006 NASA Langley Research Center 29
Grid Coarse in Radial Direction
May 8, 2006 NASA Langley Research Center 30
Grid Cause of Vorticity Lines
May 8, 2006 NASA Langley Research Center 31
Detailed PAA Flow
Analysis
Begin with Highly Complex
LSAF Jet-Pylon Nozzle
Geometries
JET3D Noise Source
Map Trends Validated
with LSAF Phased
Array Measurements
JET3D Validation of Spectra
Trend at 90 degrees
Develop Linkages of
complex flow and noise
source interactions
Three major effects to
understand:
Pylon effect
Chevron effect
PAA T-fan effect
and their interaction
PAA Analysis Process to Develop Understanding of PAA T-fan
Nozzles Flow/Noise Source Mechanisms