Hostname: page-component-7b7567c8c-wxvwp Total loading time: 0 Render date: 2026-11-15T06:34:57.175Z Has data issue: false hasContentIssue false

Weakly nonlinear development of shock cells in imperfectly expanded supersonic jets

Published online by Cambridge University Press:  28 August 2026

Zhengxuan Song
Affiliation:
School of Mechanical Engineering, Tianjin University, Tianjin 300072, PR China
Yihong Fang
Affiliation:
School of Mechanical Engineering, Tianjin University, Tianjin 300072, PR China
Zhongyu Zhang
Affiliation:
School of Mechanical Engineering, Tianjin University, Tianjin 300072, PR China
Xuesong Wu*
Affiliation:
School of Mechanical Engineering, Nantong University, Nantong 226019, PR ChinaDepartment of Mathematics, Imperial College London, 180 Queen’s Gate, London SW7 2AZ, UK
*
Corresponding author: Xuesong Wu, x.wu@ic.ac.uk

Abstract

Content of image described in text.

A weakly nonlinear theory is developed for shock cells in an imperfectly expanded axisymmetric supersonic jet. By adopting a jet profile with a finite-thickness shear layer and a high-Reynolds-number asymptotic formalism, the shock cells are treated at leading order as a superposition of a countable number of standing-wave modes. The initial amplitude of each mode at the nozzle exit is proportional to $\Delta p$, the difference between the nozzle exit pressure and the ambient pressure. Effects of viscosity and non-parallelism, associated with the slow streamwise variation of the base flow, are accounted for. The nonlinear quadratic interactions among the leading-order modes generate sum and difference modes, as well as the harmonics and mean-flow distortion. These components interact with the leading modes at the cubic level, contributing usual nonlinear effects. In addition, we have identified for the first time resonant interactions among triplets and quartets of the shock-cell modes, which follow from a simple but hitherto unnoted fact that the eigenvalues form an arithmetic sequence to a remarkable degree of accuracy. The effects of these nonlinear interactions, resonant or non-resonant, are clarified through a weakly nonlinear analysis, which exhibits rather unique features and leads to a novel system of amplitude equations coupled with a set of parabolic partial differential equations governing the mean-flow distortion. The coefficients of all the nonlinear terms in the amplitude equations are purely imaginary, which means that the non-resonant nonlinear interactions modulate directly the phase of each mode, which then influences indirectly the amplitudes of other modes through resonant coupling. The resonant interactions directly facilitate the energy transfer among different modes, as a result of which a mode may be influenced and even excited by interactions of other modes, and certain modes may undergo amplification. Through modulating the phases and amplitudes of the constituting modes, the nonlinear interactions influence the overall shock-cell features and their spacing. The predicted shock-cell characteristics and spacing are found to be in favourable agreement with experimental data. The fundamental modes and nonlinearly generated sum/difference modes exhibit considerable magnitudes in the shear layer, the possible significance of which for the emission of shock-associated noise is discussed.

Information

Type
JFM Papers
Copyright
© The Author(s), 2026. Published by Cambridge University Press
Figure 0

Figure 1.Figure 1 long description.Schematic of the shock-cell structures in an imperfectly expanded axisymmetric supersonic jet.

Figure 1

Figure 2.Base-flow quantities of an underexpanded supersonic jet with M=2.24$M=2.24$ and Md=2.0$M_d=2.0$: (a)$(a)$ thickness and central position of the shear layer, b$b$ and R0$R_0$; (b)$(b)$ streamwise velocity profiles at three x$x$ positions; (c)$(c)$ streamwise velocity U¯j$\bar {U}_{\!j}$ along the jet axis r=0$r=0$.

Figure 2

Figure 3.Figure 3 long description.Eigenvalues (a$(a$,c$c$,e$e$,g)$g)$ and eigenfunctions (b$(b$,d$d$,f$f$,h)$h)$ of the first four shock-cell modes: (a$(a$,b)$b)$ mode 1, (c$(c$,d)$d)$ mode 2, (e$(e$,f)$f)$ mode 3, (g$(g$,h)$h)$ mode 4. The eigenvalues and eigenfunctions of the vortex-sheet model are plotted using red triangles and red lines, respectively. The eigenvalues and eigenfunctions calculated by a direct method are plotted in circles.

Figure 3

Figure 4.Difference wavenumbers between two consecutive modes, Δαn=αn+1−αn$\Delta \alpha _n=\alpha _{n+1}-\alpha _n$, for the first eight modes (n=1−7$n=1-7$). The wavenumbers of the first three modes are plotted as red lines.

Figure 4

Figure 5.(a)$(a)$ Linear growth rate ϵ¯κn$\bar {\epsilon }\kappa _n$ vs x$x$, and (b)$(b)$ modulus ϵsA^n$\epsilon _s\hat {A}_n$ vs x$x$ for each mode with Δp=0.065$\Delta p=0.065$. Each modal amplitudes of the vortex-sheet model (3.15) and linear prediction (3.66) are plotted in triangles and dotted lines, respectively. Colours in (b)$(b)$ match those used for the corresponding modes in (a)$(a)$.

Figure 5

Figure 6.Figure 6 long description.Phase function Φ^n,N$\hat {\varPhi }_{n,N}$ generated by nonlinearity for each mode with Δp=0.065$\Delta p=0.065$.

Figure 6

Figure 7.A comparison between the linear and nonlinear solutions at r=0$r=0$ for the pressure of the first five modes (a$(a$–e)$e)$ and the overall pressure (f)$(f)$.

Figure 7

Figure 8.Figure 8 long description.Shock-cell pressure spectra for M=2.24$M=2.24$ and Md=2.0$M_d=2.0$ at four streamwise locations: (a$(a$,b)$b)$x=2$x=2$, (c$(c$,d)$d)$x=8$x=8$, (e, f) x=10$x=10$, and (g$(g$,h)$h)$x=12$x=12$. Left column: pressure spectra at r=0$r=0$ (centreline); right column: maximum pressure spectra for r⩾1$r\geqslant 1$ (in the shear layer).

Figure 8

Figure 9.Mean-flow distortion (a)$(a)$ϵs2u00$\epsilon _s^2u_{00}$, (b)$(b)$ϵs2θ00$\epsilon _s^2\theta _{00}$ and (c)$(c)$ϵs2ϵ¯v00$\epsilon _s^2\bar {\epsilon }v_{00}$ at three axial positions.

Figure 9

Figure 10.Figure 10 long description.Streamwise velocity (first column), temperature (second column) and radial velocity (third column) of the sum and difference modes at three axial positions: (a$(a$–c)$c)$ first harmonic, (d$(d$–f)$f)$ sum mode with m=2$m=2$ and n=1$n=1$, (g$(g$–i)$i)$ difference mode with m=2$m=2$ and n=1$n=1$, (j$(j$–l)$l)$ first fundamental mode.

Figure 10

Figure 11.Contour lines (a$a$,c$c$) and colour-filled contour plots (b$b$,d$d$) of the shock-cell pressure ps$p_s$: (a$a$,b$b$) linear solution, (c$c$,d$d$) nonlinear solution. The axial positions of shock reflection points are indicated by the red (or white) dashed lines, and the spacing of the k$k$th shock cell is denoted as sk$s_k$.

Figure 11

Figure 12.Axial distribution of the shock-cell pressure at three representative radial positions: (a)$(a)$r=0$r=0$, (b)$(b)$r=0.5$r=0.5$, (c)$(c)$r=0.9$r=0.9$ for Δp=0.065$\Delta p=0.065$. Significant small-scale structures are indicated by the red rectangles.

Figure 12

Figure 13.Figure 13 long description.Axial distribution of the pressure of (a$a$) the first mode, and (b$b$) the fifth mode at r=0$r=0$ for different Δp$\Delta p$. The linear solution is plotted using a red line.

Figure 13

Figure 14.Contour lines (a$a$,c$c$,e$e$,g$g$) and 677our-filled contour plots (b$b$,d$d$,f$f$,h$h$) of the normalised shock-cell pressure ps/Δp$p_s/\Delta p$ for different Δp$\Delta p$: (a$a$,b$b$) linear solution, (c$c$,d$d$) Δp=−0.01$\Delta p=-0.01$, (e$e$,f$f$) Δp=−0.05$\Delta p=-0.05$, and (g$g$,h$h$) Δp=−0.1$\Delta p=-0.1$. The red (or white) dashed lines indicate the axial positions of the shock reflection points.

Figure 14

Figure 15.Variation of the first shock-cell spacing s1$s_1$ with |Δp|$|\Delta p|$ for overexpanded and underexpanded jets. The fitted curves follow s1=4+245(Δp)2$s_1 = 4 + 245(\Delta p)^2$ (overexpanded) and s1=4+115(Δp)2$s_1 = 4 + 115(\Delta p)^2$ (underexpanded) for M=2.24$M = 2.24$, and s1=2.64+32(Δp)2$s_1 = 2.64 + 32(\Delta p)^2$ (overexpanded) and s1=2.64+20(Δp)2$s_1 = 2.64 + 20(\Delta p)^2$ (underexpanded) for M=1.67$M = 1.67$. The experimental data of Norum & Seiner (1982b) are marked with red upper triangles (overexpanded) and down triangles (underexpanded).

Figure 15

Table 1.Experimental data from Norum & Seiner (1982b), where the values of Δp$\Delta p$ are calculated by (4.7), and the shock-cell spacings are obtained by multiplying the peak spacings by 90%$90\,\%$.Table 1 long description.

Figure 16

Figure 16.Figure 16 long description.Schlieren images of Li et al. (2021) (left column) and contours of the shock-cell pressure predicted by the present theory (right column) for the fixed design Mach number Md=1.0$M_d=1.0$ and four different M$M$: (a$(a$,b)$b)$M=1.16$M=1.16$, (c$(c$,d)$d)$M=1.19$M=1.19$, (e$(e$,f)$f)$M=1.22$M=1.22$, (g$(g$,h)$h)$M=1.40$M=1.40$. The white dashed lines indicate the axial positions of the first three shock reflection points.

Figure 17

Table 2.Comparison of the spacings of the first three shock cells for different M$M$ at the fixed design Mach number Md=1.0$M_d=1.0$ between the present results (values outside the brackets) and the experiments of Li et al. (2021) (values inside the brackets).Table 2 long description.

Figure 18

Figure 17.Variation of the first shock-cell spacing s1$s_1$ with M$M$ in choked jets (Md=1.0$M_d=1.0$).

Figure 19

Table 3.First shock-cell spacings in the experiments of Edgington-Mitchell et al. (2021, 2022) and Nogueira et al. (2022), and the numerical simulation of Li et al. (2020) for choked jets (design Mach number Md=1.0$M_d=1.0$).Table 3 long description.

Figure 20

Figure 18.Figure 18 long description.(a)$(a)$ Contours of the axial velocity field of a choked jet (Md=1.0)$(M_d=1.0)$ with M=1.19$M=1.19$. Top: DNS data of Li et al. (2020); bottom: prediction by the present theory. The black dashed lines indicate the axial positions of the first four shock reflection points. (b)$(b)$ Axial distribution of shock-cell density along the jet centreline.