Embed Size (px)
Citation preview
4t
Computational Aeroacoustics by the Space-Time CE/SE Method
by Citing Y. LohT;dtech. Inc. and NASA Glenn Research Center
The Space-Time Conservation Element and Solution Element Method, or CE/SE Method for short, is a
newly developed numerical method for conservation laws. Despite its second order accuracy, it possesses
low dispersion errors and low dissipation. The method is robust enough to cover a wide spectrum of com-
pressible flows: from weak linear acoustic waves to strong discontinuous waves (shocks). An outstanding
feature of the CE/SE scheme is its novel, simple but effective non-reflecting boundary condition (NRBC),
which is particularly valuable for CAA (computational aeroa¢oustics).
In this seminar, the 1-D and 2-D unstructured version of the CE/SE schemes are first briefly described. Sec-
ondly, some discussions on the NRBC axe given. Then, various exmples for linear, nonlinear aeroacoustics
are presented. These include (but not limited to):
1. CAA Workshop problems (compared with exact solutions),
2. shear laver instbility, vortex roll-up (compared with linear theory)
3. jet noise -Mach radiation and jet as a Mach wave guide.
4. vortex-shock interaction, vortex-blade interaction (BVI).
5. jet noise - screech of underexpanded jet (compared to experiment ),
6. jet noise - 3-D underexpanded rectangular jet,
7. aeroacoustic feedback system - cavity noise, whistling nozzle.
At last, if time available, video animations will be shown for some of the above examples.
/
i!
/t!
This is a preprlntor reprintofapaper intendedfor presentationat aconference.Becausechangesmay be madebefore formalpublication,thisis madeavailable withthe understandingthat it willnotbe citedorreproducedwithoutthe permissiono! the author.
https://ntrs.nasa.gov/search.jsp?R=20010059234 2018-09-13T14:53:21+00:00Z
0
o
I
,i
-.<,
A
_ !I !I tl
i i ( _ )
": _ _,,L_'_: _¸_,_ _ ;L
I1
Ii
!1I!
5 ¸ )
.... . ._ _ ,.._:_/.7_;_i:_._._'_;_, '_, _:_ _._. _
©
-4_
©
©
©
I_._, _ _ _":, _,. ,
_i_%_! _I _
• - . _ - • I.-.,
C
• _ A I_¸
3 CEs ""
(quadrilaterais)
D
B
Fig. I: ba_s_ic CE/SE grid
structure
F' C' E'
time lev_.+l
A'_ B'
i i i li _ I!1
II_1 llc iEi
t,m ,evo,n....;_hg_.agon oylil_r ADB_CFA .,A'D'B'E',C'F'A' and
I
i
.÷
....."....._ ._I........_I [_,[i_._:-_!__i _ _ i! _ T t _ t t_r_!*_!:_
\©
©
m
\
,.., \
o,x
,.< \
,...\
i
I
SSg.Opi.eciJod/is+:,r,b+.+ dl,,ioz, d%,J_rv ...... , ..... e. - -
w
.ill
V
, a:'_..:?_, ,, • ;,
4.2 -:
P0.00000 ,i_,,_, i ......../i;:.'":::""
4).00005 +-..
-0.00010
-0.0001. 0 -50 0 ...... : 50 100X
• ,:,
2O_) II Print II 14 A W 1997 II fOff,pll'_:pll II
Cat.3.1, ia=O,ep=.O5,At=.6,t=100.2, cntrline (apprx. vsn.)
0.00015 iL--.GXaCt
• o.ooo_o_ CE/SE
0.00005
P0.00000
-0.00005
-0.00010
..0.00015-100
/
....I,== ...... .... I..,,.I .... ,J
-50 0 X 50 100
-. :-. -;
II
11
II
0I
II
0
. . ..._.,.: _, ., ,_ _ _:,_'U ¸,_,
• ,?.
"%>:, },:I:
/
0
, II I
0
'II! /
6ca_-Op/c_,qJod,'d_o L_a_.'di'lOOd_,_s .. 69_-_q_o_ q:_l_) _'_:% _J',:.1:_
:_i!;_'_=,_iL'-_: ii_'_
¸¸.-; i;_-, _;_!./._ i: _
o.ooo,lp
0.00005
-0.00005
:,:-_:__'_/_,'C,i'<.._7..:¢_':'_L_ r L . ....
i I
-o. _ • 0 oI., |
x
.... ';'!;;!:_: :I ' I_:.;''
..l,, II | • • • . • |
50 100
#'
(2D) It Print It 12 Jun 1997 I1 xrho.ptt rho.plt 11
Cat.3.2, Ax=Ay=l ,At=.6(CFL=.91),ep=.1 ,t=99.6 (appx. vsn.)0.00015 -
0.00010iII_0.00005
P0.00000
-0.00005
-0.00010
-o.ooo L
exact
........... CE/SE
• • • • | L • •
-50• ,,.. I • • . .. I • • • • I
0 x 50 100
_i_!i!,ii_ _,_
0
SO I.l.-OP,Sq;ed/i_'nb_'cll_i_'°[j_:_n_c;O_"l.-s,q_edLlo_i:e/'ellt'^''_I_-_
I:0oooo
0.999900 20 40 60 80 1O0X
!i,20) 1I P..rintil1"'_.'Aug 1997 IIexy100.plt nx),lbO.plt I_1
p along x=y,ep=. 1,At=.6,at t= lO0(Cat.4.1 )
! ooo o!-t I ............ exact
1.00000_
iiiiii ''O0
_._,_:i_>;_ ..
..... ,_" _ :
.; :.. ,.... _.-_._.'_.}_':_.:. ; , .
• t: [-'_ " q i' ' "
in the computation.
0.010.01
3i-- 5
L)
1
2 i- "'" real part, of exact sol.CE/SE solution at y=O
t_
1:--,
&
-1 -
'2,
1
2
/
iI
,,,%
0.0 0.2 0.4 0.6 0.8 1.0
..
3
"_s"
aa., imaginary part of exact s61.
_ CE/SE solution at y=3 :
l
0.0 0.2 0.4 0.6 0.8 1.0
....=,,.,.,,,.2: Pressure c!Jfference ac_-o,_s.<_-k_,,,,,..ui_._.c_;_,u,_surfac:e co_Tnpkrec" w_r4-,,"'
:.i_: c:._.c_.bu bt)lttbluli,
0.030 t ............. _.
o.o2si j
' t0.005 ! Exact solutiono.oooi _""
0.0 0.2 0.4 0.6
(a) x=-20.8 1.0
o.o3u:": , _ ....
// \ I
0 020// .
o.oIo_ _:J
. _ Exact solution :.P
0.000: .............
0.0 0.2 0.4 0.6 0.8 1.0
(b) ×---3
_.i:.,.,,,-,.:_:The sound intensity at x = -2 and x = 3 compared with
,,,it expect aolution.
2
_i'..-.!_.__:!_i __ _;_.'_;_._
::]
.i :.i_!_̧ :i _
0
0
©
©
Q
00
0
)
?
. .. _-__ .:.T_._".-_ _:,_,_,__ :-_.,_:_ _- ..... ,.,_ : ..........
..... .........:":;"_'"t • ":-..... '::"..... :!:_gg;()p/_ledhsenbeJ/dlllOOds/Jsn/_j_t_qJed qOl_l:s._vqGEl3
, : i.
0.020
-o_
0.010
o.ooo
i
P.
-8.0
-14.0
r
I I
-20.0
-_'_._ " o.i_ ' o.;o " o._ ' _.;o
Figure: Power sp_tr_ _ • = 150 and 250 for natural (unforced) case; coarse grid.
_'_" ,_._".tta. _:7:'#_:_'_>_'_¢_ '_'" ""' ' :'_: .....
Figure: Power spectra at x = 050, 100, 150, and 250 with forcing at the most unstable frequency
according to linear theory; coarse grid.
[
, t
- "_.oo 0.25 0.50 0.75 .
/.
Figure: Power spectra at x = 050, 100, 150, and 250 with forcing at most unstable frequency
according to linear theory; finer grid.
.H
o.oo olin o.so o.Ts x.ooI'I
Figure: Transverse mode shape at x = 100 with forcing at most unstable frequency acording to
linear theory, finer grid. Squares: total urms; circles: urrns at forcing frequency; solid line: linear
cigenfunction (modulus).
Y
9.0
-3.0
m
--g.O | I __ l I ! I I I-1.2 -0.6 0.0 0.6 1.2
Figure: Transverse phase variation at x = 100 with forcing at most unstable frequency acording to
linear theory, finer grid. Circles: urm_ at forcing frequency; solid line: linear eigenfunction (phase).
............................ _- _210
!
Figure: Streamwise evolution of di.sturbance amplitude with forcing at the most unstable frequency
according to linear theory. Squares: total urine, coarse grid; triangles: total _rm$, finer grid; circles:
ur_n, at forcing frequency, coarse grid; crosses: Urm, a_ forcing frequency, finer grid; solid line:
linear growth.
45.0 ,-
30.0
15.0
0.0 ,, I , ,I
O. 100. 200. 300.
]"
Figure: Streamwise evolution of disturbance phase with forcing at the most unstable frequency
according to linear theory. Circles: coarse grid; crosses: finer grid; solid line: linear result.
',,! :_,_ :_-,:
"_ ' '_Q, _! _ t _ ' _,{_'f_'_ : f _"_ S!{%,,_ _ __,_"l_\ _ '_{_ i,_, __'7._ _{_
• x _: . k, L. • ,:
• _,_'_i_/ i _'.... ,._ _i_._{_ _i._i_
i
3
"iwC)
- (I)
0 5"0 0')XI'_ l:_0
:Z} 0
II (.Q0_.L 00"I --"
I
(.-"0
II
c..C)C_
0
X
0c_
C)..
-0(I)
c-
B-(I)C).
i.
_C (-"0
C_ Q,D 0
DL--_. •
1
_, ..._.-.,_-_._
\
\ ! _.-
• ",, i \
\
\\,
\
\
\
-\t
9_gg-op/_.qJed/'isenbeJ/dl/ioocl_IJ_nl_:g_ 1_q_t_d_._i.__.A_Ii_,',_-_
i
<\
\
\
\:\"_"_'"X.,
%r© _,
'\
%._
-\.
\
-.+_._
i
_k
g
ii
'Iii
.j
i
/
/
_4r..;
..>
t
C)e_;ty r-el;el _."./Pl = 1,2,'7= 18/us
Demt_it_, r, eli,:f P''Pl = I, 27 • = 9e ,/-:_-
Ycoi
:,.i_._:+Yt
..H+_:ki:...:...,_.
+y!!::_" .....'_,.,'.";z_,:,.
1
-'_ (1)
::_ CO
",,4_
_ o
il
0C_
4_0
0
00
7_
\
I
\\
/
\
, \\I
"%
N
\\,
\
\,
\
i _
i -
ql
-\
\
\
\
\ " i
-.;: . \
:Y_Z ,_, "l
\
I /i
_Z" ;._. {'_'_, _",
_;_.":_"A_._I
y;:.
l
959.0P. _,qJE,,_/isc.,rL,e.UOi iood_,._Sh.9_9-1_qJ_-'C-4:,I',,,uc_-:c-._,G:"
<!
C)0
0
CD
0
<0
(I)X
COGO
0
U)
0C')
/
E
/
i(Z) .
//
/i/\
\
k
4_
/
/._ .._...........
Ii_:_•_._,!!--,
\
i
I /I ,f
,i
I_t
I
/J
/t/,\
\
•,r •
I
. t
/
I /.J(r, :,:.']'.L":,,.
\s,
\
%
J
t/
\
•,%/.% j'.s
_"'I \\
_il _ . • ..
iPO _
/
//
%
\
/Y
/
.,;:.."-" iA',
- .)
\
,¢
I
//
/
• \ • t
tt
,._,, • ,
&_-', ..,_..,_ .... ,,. .... , .... _.. , . .
x_P__ as
pet 4/_a r ( _ft r
\ /\ /'. /\ / \\-/\ / \ / ,.,/ ',,/ '\\/ ,,/\/ \/ ,1 v ,_, ./ y ,/
. X 7,. .X A. /',, /5, ,,,, ,,',./',. ,,,,\ / \ / \ / ",._/\, / \ / ",,_,",.,,/"..,/ ,,/
• ,,i
Cr'o_;,s-haYc.hed Mac A N=v_
'm,b,lO
11W
v_
II
_J%
l!
qk_
Pil__-,2 , S'_= o.o 7, b"3c,* / _t_ .
L
• ,_JI
o
(
1
t'\
shear layer _.
_I--'--" .'".- ._-. " -,'.',-_I"-_-_r; _ " _ .- '' ,;,___';X_."""
:_',_,-_ ___N_:,
Li _ - _ _ ..
shear layer -/
(!
Ivl.._.j=t.2:3Dx1D.,300xi00 grid, St. =2. dt=.OO15,ee=.O0005:
showing cross-hatching Mach wave sys'_em.
4mmll
CD
a_
C_a_
C_
Oral
Omc
q_
_L
i
it
!!
!
m
I
I
i!!!IIII1II
C2
Figure 1: .Numerical Schlieren b v CE/SE (snapshot)
Fig3_re 2: Experimental Schlioren (courlesy of J. Panda l
LoII
c_
xct_
3
CD
m
Q__D
_Dx
QI_D
QI
mw
CQ
screech tone waves (wavelength=l.6D)
,i_ ynam
nozzlell
isobar and numerical Schlieren snapshotsat time step=212800, underexpanded a×i-sym. jet, M_j=I .19.
\. / x / °,-¢_,t-\ / \ l "
, \ ! 0_a
\ f \ I =_\ t \ /
\ / \ tx I \ I
Raman'sexperiment
numerical Schlierens
Raman's
experiment
numerical Schlierens
isobars,--narrow Si-de mid sect. plane
isobars, wide side mid sect. plane
u-contours, na-rro--W_i-d-e-mtd _1_ b-he
Y
(- , _.• "- '._'-.-.., . ;- .
u-contours, W i _-S !_le..LTl_'al-ptarle
shock-cells
p iso-surfacefrom the wide side
low level: p=.274, high level: p=.2792
p iso-surfacefrom the narrow side
non_,_,n_r ou_ro_ _s "t_cS_--_ _._ _-- _
A
_mpzngzng
waves
D M=0.8
fcavl
B
H=o 8 vorti _mp_ng_ng
: waves
rfla t slab
Fig. 2: Sketch
(B)of
gap
(A) cavity
nolse
nolse,
I
"X
Deutsches Zentrumf6r Luft- und Raumfahrt e.V
German
Aerospace Center
DLR Institute of Design Aerodynamics
Postfach 3267, D-38022 Braunschweig, Germany
Dr. Ching Y. Loh
Taitech Inc.NASA Glenn Research Center21000 Brook Park RoadMS 5-11
Cleveland, OH 44135
U.S.A.
Cluster Aerodynamicsand Flow Technology
Your letter
Your reference
Our reference
YourcorrespondentTelephone+49 5 31 2 95-
Telefax+49 5 31 2 95-E-mail
Braunschweig,
Institut of
Design Aerodynamics
Prof. Dr.-Ing. [email protected]/11/00
Invitation for the second SWING Aeroacoustics Workshop 6-7 Oct. 2000 at OLR
Dear Dr. Loh,
In the framework of our national German CAA-project SWING we are organizing a workshop at DLRin Braunschweig in cooperation with the University of Technology Dresden. I would very much liketo invite you to this second SWING workshop which will be held on 6-7 October this year. SWING(Simulation of Wing-flow Noise Generation) is an initiative geared towards the simulation of airframenoise sources at high lift devices using modern numerical CAA methods. The project partnershipconsists of DLR's Institute of Design Aerodynamics, the Aerodynamic Institute Aachen (TechnicalUniversity of Aachen), the Institute of Aerodynamics and Gasdynamics (University Stuttgart) as wellas the Institute of Acoustics and Speech Communication (University of Technology Dresden).
The aim of the workshop is to exchange knowledge on modern topics of aeroacoustics with specialemphasis on numerics and to establish/deepen international connections and cooperation amongresearch groups active in the respective field. The first SWING workshop, held in the fall of last yearin Dresden, led to intensive and fruitful scientific discussions; and this is why we would like tocontinue this successful endeavor. During the 1.5 days of the meeting there will be half-hourpresentations by the SWING project partners and by invited speakers from abroad and Germany.
It is my particular pleasure to invite you to this workshop to give a presentation about your recentresearch work relating to the workshop topic. Fortunately, the expenses connected with your visitand stay in Braunschweig can be covered since we are able to refund you with a bulk sum of DM
3500,- ( = $1500,-).
We thank you for your willingness to present the space-time conservation method for aeroacousticsat the meeting as already agreed upon with Dr. Jan Delfs, who is in charge of organizing theworkshop. Looking forward to seeing you in Braunschweig.
Yours sincere_
H_K_!n//e_"s'_attachment: proceedings of first SWING workshop
Lilienthalplatz 70-._8lOa n_au¢,schweig
Germany
