Anisotropic Pressure and Acceleration Spectra in Shear Flow Yoshiyuki Tsuji Nagoya University Japan Acknowledgement : Useful discussions and advices were given by Prof. Y. Kaneda Objective T. Ishihara, K.Yoshida, and Y.Kaneda, Anisotropic Velocity Correlation Spectrum at Small Scales in Homogeneous Turbulent Shear Flow, Phys. Rev., Letter, vol.88,154501,(2002) Shear effect on inertial-range velocity statistics are directly investigated . This idea is applied to the pressure field in the uniform shear flow, and the shear effect on pressure and pressure gradient (acceleration) is studied experimentally up to the Reynolds number based on Taylor micro scale is 800. 2. Pressure Measurements Pressure measurement Kolmogorov length scale is 0.19mm for R 700 . Φ=0.08mm Φ=0.15mm 2.0 Φ=0.3mm Φ=0.5mm 0.4mm d 12mm l δ Microphone Microphone: 2 10 4 ~p 3.2 10 3 [Pa] 2 101 f 7.0 10 4 [Hz] 1/8 inch d 3.2 [mm] pressure measurement inside the boundary layer Probability density functions Probability density 100 10-1 10-2 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 0.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 p/ p/ 10-3 10-4 :EXP(R =200) :DNS(R =164) :EXP(R =320) :DNS(R =283) 10-5 -12.0 -8.0 -4.0 0.0 4.0 p/ -12.0 -8.0 -4.0 0.0 4.0 p/ DNS: Kaneda & Ishihara Nearly homogeneous isotropic field. Kolmogorov constant Pressure Spectrum R 600 R>600 (DNS:Gotoh,2001) Nearly homogeneous isotropic field. Pressure measurement in Boundary layer Epp/[ 2 2 u ] 104 slope=-1.0 103 slope=-1.2 102 101 100 10-1 10-2 10-3 + y =200 R=5875,7420,8925,10515,12070,15205 :Abe et al. (R =2066) 10-4 10-3 10-2 f/u2 10-1 100 -7/3 power-law is not observed in the overlap region of smooth-wall boundary layer even if the Reynolds number is very high. Pressure spectrum in the boundary layer 2. Experiment Experiments: Driving Mixing Layer y Mixing layer centerline Transition region d=350mm Potential Core x/d~5 x/d Nozzle exit Mixing layer centerline L=700mm In this region, flow reversals are unlikely and large yaw angles by the flow are infrequent. x Driving Mixing Layer :x/d=1 :x/d=2 :x/d=3 :x/d=4 :x/d=5 x/d=5 U/UJ 1.0 0.5 0.0 -0.2 x/d=4 -0.1 urms/UJ y 0.2 0.3 0.2 x/d=2 Nozzle exit 0.1 y/(x-x 0) x/d=3 x/d=1 0.0 :x/d=1 :x/d=2 :x/d=3 :x/d=4 :x/d=5 0.1 0.0 -0.2 -0.1 0.0 0.1 y/(x-x 0) Nearly homogeneous shear flow. 0.2 0.3 Reynolds number & Shear parameter Reynolds number S* 10-1 R u ' u x / u ' 2 2 10-2 Shear parameter 102 S S k * 103 R 1/ 2 1 20 1 S flow dU dyS k R R 1 Simple uniform shear A 3 * Driving mixing layer is close to the simple uniform shear flow. 3. Theoretical formula Shear effect on velocity fluctuation According to the formula presented by Ishihara, Yoshida and Kaneda PRL(vol.88,154501,2002), velocity spectrum is defined by Qij (k , t ) 1 (2 )3 ik r dr ui x r , t u j x, t e 0 Qij (k , t ) Qij (k , t ) Qij (k , t ) S12 U y :Simple mean shear s K1 S2 /:independent of wave number k 3 11/ 3 0 Qij (k ) k Pij (k ) Isotropic part (K41) 4 u :dependent of wave number k 0 N Modification due to the existence k 1 of mean shear. :characteristic eddy size Qij (k , t ) Cijmn (k )P(k )S mn Cijmnkl (k ) R(k ) S mn S kl Anisotropic part u :characteristic velocity scale u 1/ 3 1/ 3 13/ 3 k k Cijmn (k ) A Pim (k ) Pjn (k ) Pin (k ) Pjm (k ) k BPij (k ) m 2 n S ij d U i dx j k 2 for large wave numbers N S Pij ij ki k j k Shear effect on velocity fluctuation Velocity spectrum is obtained by the summation with respect to k over a spherical shell with radius k . 0 Eij (k ) Qij (k , t ) Q ij (k , t ) Qij (k , t ) k k k k ˆ ˆ E (k ) k a kbQij (k , t ) ab ij k k Isotropic part (K41) Anisotropic part E12 (k1 ) E11 (k1 ) k k In usual experiments, one-dimensional spectrum is obtained. 18 5 / 3 K 0 2 / 3 k1 55 4 18 5 / 3 K 0 2 / 3k1 3 55 is proportional to mean shear 36 33 A 7 B 1/ 3 S12k17 / 3 1729 432 2 A B 1/ 3 S12k17 / 3 E (k1 ) 1729 12 11 E22 (k1 ) u1u2 E12 (k1 )dk1 du1 du1 12 E11 (k1 )dk1 dx1 dx2 Isotropic velocity spectrum 100 E11(k1 ) , E22(k1) 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10 -8 10-9 -1 10 R=710 :E 11(k1) :E 22(k1) 100 101 102 103 104 k1 Isotropic part (K41) 18 5 / 3 E11 (k1 ) K 0 2 / 3 k1 55 4 18 5 / 3 E22 (k1 ) K 0 2 / 3k1 3 55 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 600<R<700 10-12 10-13 10-14 -1 10 100 101 E1112(k1)/[1/3S12] E12(k1)/[1/3S12] Anisotropic velocity spectrum A 0.17 102 k1 103 104 100 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 -13 B10 -140.45600<R <700 10 10-15 10-16 -1 10 100 101 102 k1 103 104 Anisotropic part is proportional to mean shear even if S is changed. 12 E12 (k1 ) 36 33 A 7 B 1/ 3 S12k17 / 3 E1112 (k1 ) 432 2 A B 1/ 3 S12k17 / 3 1729 1729 Shear effect on pressure According to the formula presented by Ishihara, Yoshioda and Kaneda PRL(vol.88,154501,2002), pressure spectrum is defined by Q p (k , t ) 1 (2 ) 3 ik r dr px r , t px, t e 0 Qp (k , t ) Qp (k , t ) Qp (k , t ) Isotropic part (K41) Q p (k ) K p 0 Anisotropic part 4/3 k 13 / 3 kk Q p (k ) 1 22 k 5 S12 k S12 U y :Simple mean shear Modification due to the existence of mean shear. Qp (k , t ) Cmn (k ) P(k ) Smn Cijkl (k ) R(k ) Sij Skl C ijkl (k ) :4th order isotropic tensor C mn (k ) :2nd order isotropic tensor Shear effect on pressure spectrum Pressure spectrum is obtained by the summation with respect to k over a spherical shell with radius k . 0 E pp (k ) Q p (k , t ) Q p (k , t ) Q p (k , t ) k k k k k k Isotropic part (K41) E pp (k1 ) x2 7 7 / 3 K P 4 / 3 k1 6 x1 Anisotropic part E pp (k1 ) 0 k1 9 / 3 S12 Ck1 11/ 3 S12 S12 0 Shear effect on pressure spectrum appears in the second order of S12 Shear effect on pressure spectrum Pressure spectrum is obtained by the summation with respect to k over a spherical shell with radius k . 0 E pp (k ) Q p (k , t ) Q p (k , t ) Q p (k , t ) k k k k k k x2 x2 Isotropic part (K41) E pp (k1 , x2 ) x2 7/3 2C1 k1x2 4/3 7 / 3 K P (k1x2 ) (5 / 3) 2 7/6 7 / 6 (k1x2 ) Anisotropic part E pp (k1 , x2 ) x2 3 5/ 2 2 C d k x 3 1 1 2 S12 (k1x2 ) 5 / 2 (k1x2 ) (5 / 3) d (k1x2 ) 2 x1 Shear effect on pressure spectrum IYK formula is well satisfied in this experiment. 108 3 106 105 Epp(k1,x2)/x2 Epp(k1,x2)/x27/3 107 104 103 102 101 100 10-1 10-2 10-3 -3 10 10-2 10-1 100 k1x2 Isotropic part (K41) 101 109 108 107 106 105 104 103 102 101 100 10-1 10-2 10-3 10-4 -3 10 2.45 10-2 10-1 100 101 k1x2 Anisotropic part 5/ 2 7/6 E ( k , x ) 2 C d k x pp 1 2 3 1 1 2 E pp (k1 , x2 ) 2C1 k1x2 4/3 7 / 3 S ( k x ) ( k x ) 5/ 2 1 2 3 K P (k1x2 ) 12 71/ 6 (k21x2)(5 / 3) d (k x ) 2 7/3 x 1 2 (5 / 32 ) 2 x2 Shear effect on velocity&pressure According to the formula presented by Ishihara, Yoshioda and Kaneda PRL(vol.88,154501,2002), velocity&pressure spectrum is defined by Ri (k , t ) 1 (2 )3 ik r dr px r , t ui x, t e 0 Ri (k , t ) Ri (k , t ) Ri (k , t ) Isotropic part (K41) Ri (k ) 0 0 Anisotropic part Ri (k , t ) Cimn (k )r (k )Smn Cimnkl (k )s(k )Smn Skl kk k k k k Cimn (k ) a i m3 n b im n mn i ni m kˆ kˆ kˆ kˆ Cimnkl (k ) :5th order isotropic tensor Shear effect on velocity&pressure spectrum Pressure-velocity spectrum is obtained by the summation with respect to k over a spherical shell with radius k . 0 E pui (k ) Ri (k , t ) R i (k , t ) Ri (k , t ) k k k k k k ˆ E (k ) k j Ri (k , t ) j pui du1 p k1E1pu1 (k1 )dk1 dx1 0 k k Anisotropic part (i 1) E pu1 (k1 ) 0 (i 2) du 2 p 1 k1E pu (k1 )dk1 1 dx2 0 (i 1, j 1) E1pu1 (k1 ) 0 (i 2, j 1) 6 2/3 9 2/3 18 9 8 / 3 8 / 3 2 E pu2 (k1 ) a b S12k1 E pu1 (k1 ) a b S12k1 11 48 187 140 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 (i 2) Epu12(k1)/[2/3S12] Epu2(k1)/[2/3S12] Shear effect on velocity&pressure spectrum 101 102 103 k1 104 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 0 10 101 102 103 k1 (i 2, j 1) 104 a 0.1 b 0.03 6 2/3 9 2/3 18 9 8 / 3 8 / 3 2 E pu2 (k1 ) a b S12k1 E pu1 (k1 ) a b S12k1 11 48 187 140 Isotropic velocity spectrum 100 E11(k1 ) , E22(k1) 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10 -8 10-9 -1 10 R=710 :E 11(k1) :E 22(k1) 100 101 102 103 104 k1 Isotropic part (K41) 18 5 / 3 E11 (k1 ) K 0 2 / 3 k1 55 4 18 5 / 3 E22 (k1 ) K 0 2 / 3k1 3 55 Acceleration In a usual notation, pressure relates to acceleration vector ; 2 a p u p 1 p ai xi dp dp (t ) dt dp (t ) 1 dx dt dx dt Uc 1 p aj x j i, j 1,2,3 U c Local mean velocity dp(t ) nmax 2f n an sin 2f n t 2f n bn cos2f n t dt n 0 2f n max 1 / k Shear effect on acceleration Similar discussion is possible in case of acceleration i p xi . E i j (k ) ki k j k k k2 ki k j 0 ki k j Q p (k , t ) 2 Q p (k , t ) 2 Q p (k , t ) k k k k k k Isotropic part (K41) E11 (k1 ) A1 4 / 3k1 1/ 3 E 2 2 (k1 ) A2 4 / 3k1 1/ 3 0 k1 3 / 3 0 k1 S12 C1k1 3 / 3 5 / 3 S12 C2k1 Anisotropic part S12 S12 5 / 3 S12 S12 2 , there is no As far as looking for the variance of 1 5and 3 / 3 /3 4 / 3 1/ 3 E (k1 ) 0effect k1by shear. C3k1 S12 0 k1 S12 S12 significant 1 2 11 E (k1 )dk1 0 1 1 2 2 E (k1 )dk1 0 2 2 1 2 E (k1 )dk1 0 1 2 Kolmogorov scaling for acceleration Following the Kolmogorov’s idea, acceleration is scaled by energy dissipation and kinematic viscosity, and the constant a 0 becomes universal. 2 a p u p 1 p ai xi ai a j a 0 1 p aj x j i, j 1,2,3 1 / 2 ij 2/3 a 0 :Universal Constant Kolmogorov scaling for acceleration a1 a1 a 0 a0 101 :(2.5*R0.25 +0.08*R0.11 )/3.0 100 S 0 * 102 :Mixing layer 103 R :DNS(Vedula&Yeung) :DNS(Gotoh&Fukayama) a 0 is not constant but increases as Reynolds number increases. There is no significant difference between S * 0 and S * 0 1 / 2 2/3 S* 0 Summary : pressure • In a simple shear flow, shear effect doe not appear clearly in a single-point statistics. x2 Anisotropic part E pp (k1 ) 0 k1 9 / 3 S12 Ck1 11/ 3 S12 S12 x1 Shear effect on pressure spectrum appears in the second order of S12 • Shear effect can be evaluated by two-point statistics. x2 Anisotropic part 2.45 E pp (k1 , x2 ) x2 3 5/ 2 2 C d k x 3 1 1 2 S12 (k1x2 ) 5 / 2 (k1x2 ) (5 / 3) d (k1x2 ) 2 x2 x1 Summary : pressure-velocity correlation • In a simple shear flow, shear effect on pressure velocity correlation is evaluated by the relation. Anisotropic part (i 2) 6 18 8 / 3 E pu2 (k1 ) a b 2 / 3 S12k1 11 187 (i 2, j 1) 9 2/3 9 8 / 3 2 E pu ( k ) a b S k 1 12 1 1 48 140 a 0.1 b 0.03 pu2 E pu2 (k1 )dk1 0 du 2 p 1 k1E pu (k1 )dk1 1 dx2 0 Summary : Acceleration • In a simple shear flow, shear effect appears on the correlation between 1 and 2 . ai Anisotropic part E11 (k1 ) 0 k1 3 / 3 E 2 2 (k1 ) 0 k1 E1 2 (k1 ) C3k1 S12 C1k1 3 / 3 3 / 3 5 / 3 S12 C2k1 S12 S12 5 / 3 S12 0 k1 5 / 3 1 p xi S12 S12 S12 S12 • The constant a0 defined by Kolmogorov scaling of acceleration variance is not affected clearly by shear. ai a j a 0 1 / 2 ij 2/3 Frozen Flow Hypothesis for Pressure Wall pressure spectrum Frozen flow hypothesis DNS result by H. Abe for Channel Flow Wall pressure spectra Probability density function of acceleration Mixing layer 100 probability 10-1 10-2 10-3 10-4 10-5 :La Porta et al. 10 -6 -20.0 -10.0 0.0 (dp/dx)/ * small R large S 10.0 20.0 Pressure measurement in cylinder wake ~ /U / d ω 3 vorticity ~ P [ Pa] pressure Q / U / d 2 Second invariance of velocity gradient tensor 100 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 100 10-1 slope=-1/3 10-2 slope=-7/3 Spectrum Spectrum Spectra of pressure and acceleration Inertial range 10-3 10-4 Inertial range 10-5 10-6 (a) pressure 10-7 10-8 100 101 102 103 frequency [Hz] E pp ( f ) f 7 / 3 104 (b) acceleration 100 101 102 103 frequency [Hz] Epp ( f ) f 1 / 3 104