Processing math: 10%

有攻角圆锥边界层横流失稳分析

刘姝怡, 陈曦, 万兵兵, 陈坚强, 黄刚雷

刘姝怡, 陈曦, 万兵兵, 等. 有攻角圆锥边界层横流失稳分析[J]. 空气动力学学报, 2025, 43(2): 86−95. DOI: 10.7638/kqdlxxb-2024.0087
引用本文: 刘姝怡, 陈曦, 万兵兵, 等. 有攻角圆锥边界层横流失稳分析[J]. 空气动力学学报, 2025, 43(2): 86−95. DOI: 10.7638/kqdlxxb-2024.0087
LIU S Y, CHEN X, WAN B B, et al. Crossflow instability analysis of boundary layer on a yawed cone[J]. Acta Aerodynamica Sinica, 2025, 43(2): 86−95. DOI: 10.7638/kqdlxxb-2024.0087
Citation: LIU S Y, CHEN X, WAN B B, et al. Crossflow instability analysis of boundary layer on a yawed cone[J]. Acta Aerodynamica Sinica, 2025, 43(2): 86−95. DOI: 10.7638/kqdlxxb-2024.0087

有攻角圆锥边界层横流失稳分析

基金项目: 国家自然科学基金(12372225,12202477)
详细信息
    作者简介:

    刘姝怡(1994—),女,云南宣威人,助理研究员,研究方向:边界层转捩. E-mail:LSYNWPU@126.com

    通讯作者:

    黄刚雷*,助理研究员,研究方向:空气动力学. E-mail:huangganglei2021@163.com

  • 中图分类号: O35;V211.3

Crossflow instability analysis of boundary layer on a yawed cone

  • 摘要:

    横流失稳是复杂三维边界层转捩的主要机制之一。针对马赫数6条件下常规风洞实验工况的6°攻角尖锥边界层,利用二维全局稳定性分析( two-dimensional global stability analysis, BiGlobal)方法和面推进抛物化稳定性方程(three-dimensional parabolized stability equations, PSE3D)从多维稳定性分析的角度开展了横流失稳分析,并与一维线性稳定性理论(linear stability theory, LST)和直接数值模拟(direct numerical simulation, DNS)结果进行对比。结果显示,横流模态主要分布在背风面,但随着幅值的增长,模态扰动仍可显著影响迎风面区域;非定常横流模态比准定常横流模态更不稳定,且波角更小;相较于BiGlobal,基于PSE3D得到的N值略低,但两者主频相近,表明非平行效应削弱了横流扰动的增长,但并不改变扰动主频;基于LST的N值远大于全局稳定性分析得到的N值,一方面体现了两种方法对扰动增长定义的差异,另一方面也反映了三维效应的影响,因此不同的稳定性分析方法对确定转捩的N值影响显著。

    Abstract:

    The crossflow instability is one of the main transition mechanisms in complex three-dimensional boundary layers, though it is still far from being fully understood. In the present study, the two-dimensional global stability analysis (BiGlobal) and plane-marching three-dimensional parabolized stability equations (PSE3D) were employed for the crossflow instability in the boundary layer of a sharp cone at 6° angle of attack under the Mach 6 wind tunnel experimental conditions. The results were compared with those of the one-dimensional linear stability theory (LST) and direct numerical simulation (DNS). It is shown that the crossflow mode is mainly distributed on the leeward side, yet can still remarkably affect the windward side as its amplitude increases downstream. The most unstable travelling crossflow mode is more amplified and has a smaller wave angle than the stationary crossflow mode. PSE3D obtains a slightly lower N factor than that obtained by BiGlobal, while the two have similar peak frequencies, indicating that the non-parallel effect weakens the growth of crossflow disturbance, without changing the peak frequency of disturbance. The N factor obtained by LST is significantly larger than those from global stability analyses, reflecting the difference in the definition of the growth rate between those two methods as well as the influence of three-dimensional effects. Therefore, different stability analysis methods have a significant impact on determining the transition N factor.

  • 精准预测边界层转捩对飞行器热防护设计有重要意义[1-3]。在飞行状态下,边界层通常具有强三维特性。展向压力梯度会使边界层内出现一个与外部无黏流线相垂直的速度分量,即横流。横流不稳定性属于无黏拐点失稳,也是三维边界层失稳的主导机制之一。横流扰动既可以是定常的也可以是非定常的,二者谁占主导地位取决于来流和壁面条件。在马赫数6条件下的风洞实验中发现,噪声和静音环境下同时存在定常横流涡和非定常横流波[4]。在噪声环境下,定常横流涡容易受噪声干扰难以被观察到[5-6]。然而表面粗糙元的存在会增强定常横流涡,并抑制非定常横流波幅值[7-8]。此外,自由流中的慢声波可经同步机制激发非定常横流波[9]

    有攻角圆锥是研究边界层横流失稳的重要模型。圆锥迎风面和背风面不同的激波强度形成周向压力梯度,从而引起横流。美国和澳大利亚联合执行的HIFiRE-1项目中包括了圆锥转捩飞行试验[10-11],发现在大攻角时,背风面横流不稳定性主导转捩[13]。国内有攻角锥飞行试验(MF-1)在小攻角情况下获得了0~20 kHz的低频信号, Wan等[12]结合理论分析,对比信号相速度和波角等特征,推测该信号为非定常横流波。

    目前,国内外已针对有攻角圆锥边界层进行了一系列风洞实验[7,13-14],发现非定常横流波的频率在20~50 kHz之间,相速度约为0.1~0.5倍边界层外缘速度[7],且随频率的增大而增大,而定常横流涡的展向波数在40~80之间,且会随着雷诺数的增加而增加[15-16]。Craig等[17]和Saric等 [18]首次采用热线风速仪在马赫数6静音风洞中对定常横流涡的空间结构进行了定量测量。Dong 等[19]对有攻角圆锥转捩过程进行了直接数值模拟(direct numerical simulation, DNS),在背风面获得了10~50 kHz的非定常横流波,还发现壁面温度升高,横流涡的尺度也会随之增大[20]

    目前对有攻角锥横流不稳定性的理论分析主要集中在线性演化阶段以及幅值演化到一定程度后的非线性饱和阶段[21]。早期,主要采用一维线性稳定性理论(linear stability theory, LST)对有攻角圆锥进行分析,得到的转捩前缘在侧面凸出。LST需要预先确定横流扰动的传播路径和沿路径的展向波数[25-26],而全局稳定性分析方法无需预设扰动传播方向和展向波数,且全局模态均位于相同截面内,便于开展非线性相互作用研究。同时,全局模态为展向波包型分布,可作为函数基来重构真实扰动。因此,在转捩机理研究方面,全局稳定性分析更适用于展向和法向非均匀的三维流场[28]。但是由于横流模态通常具有展向分布区域广、展向波长短的特征,采用多维稳定性理论分析需要大量的内存[29]。Paredes等[30]首次将二维全局稳定性分析(two-dimensional global stability analysis, BiGlobal)方法应用在椭圆锥边界层上,得到的非定常横流模态几乎遍布整个模型表面,且展向波长较小。Lakebrink 等[31]通过对比一维稳定性分析方法和BiGlobal得到的非定常横流模态增长率,发现二者差异明显,并指出在三维边界层转捩问题分析中展向变化不容忽视。Liu等[32]采用BiGlobal得到了有攻角圆锥非定常横流模态相对完整的特征谱随频率和流向位置的变化规律,发现单个频率存在多个横流模态,并通过形状函数得到了横流模态展向波数分布。

    BiGlobal忽略了流向非平行性的影响,而面推进抛物化稳定性方程(three-dimensional parabolized stability equations, PSE3D)考虑了流向的非平行性影响,更适合研究扰动的空间演化特征。PSE3D求解的是边值问题,需要BiGlobal提供入口条件。Chen等[33]采用PSE3D对升力体模型的横流失稳区进行了分析,并将结果与DNS得到的横流扰动增长率进行了对比,相较于BiGlobal,PSE3D能更好地预测模态流向演化情况。目前,PSE3D已成功应用于有攻角锥迎风面第二模态[34]和背风中心线流向涡[35-36]失稳分析中。本文将进一步利用PSE3D研究马赫数6常规风洞实验工况下,6°攻角尖锥边界层横流全局模态的空间演化特征,并与LST、BiGlobal和DNS结果进行对比,期望揭示圆锥边界层横流空间演化规律,探索基于PSE3D的转捩阵面预测可行性。

    本文的研究模型为400 mm长的尖锥,头部半径为0.05 mm,带6°攻角,头部半锥角为 7°。研究工况为马赫数6条件下的风洞实验工况,来流单位雷诺数Re=107/m,来流温度T=48 K,模型壁面温度Tw=300 K。

    基本流采用李新亮等发展的OpenCFD计算流体力学软件[37]求解。由于尖锥的对称性,本文只考虑半模。求解分为两步,首先,求解带头部的流场,流向、法向和周向网格点数量分别为291、301和301;然后截断头部,将上一步计算得到的定常流场对应截面作为入口和上边界条件,出口处流向网格逐渐变稀疏以减弱扰动在边界处的反射。壁面为无滑移等温壁面(Tw=300 K),流向、法向和周向网格点数量分别为1000、190和800。对流项和黏性项分别采用五阶迎风差分格式和六阶中心差分格式。

    图1展示了不同流向站位归一化后的流向速度ˉu和壁面压力ˉp云图。可以看出,由于攻角效应,圆锥侧面存在明显的压力梯度,使流线逐渐向背风中心线汇聚,形成蘑菇状的流向涡。流向涡中不稳定模态主要有两类:位于近壁剪切区的内模态和位于涡顶部剪切区的外模态[36,38-39]

    图  1  计算域和基本流
    Figure  1.  Computational domain and baseflow

    本节将简要介绍LST、BiGlobal和PSE3D三种理论方法。流场瞬时量{\boldsymbol{q}} = \left( {u,v,p,T,w} \right) 可以分解为基本量\bar {\boldsymbol{q}}和扰动量{\boldsymbol{q}}'

    {\boldsymbol{q}}\left( {\xi ,\eta ,\zeta ,t} \right) = \bar {\boldsymbol{q}}\left( {\xi ,\eta ,\zeta } \right) + {\boldsymbol{q}}'\left( {\xi ,\eta ,\zeta ,t} \right) (1)

    其中\xi \eta \zeta 对应图1中流向、法向和展向贴体坐标,uvw为对应速度分量。

    LST假设只有法向是非均匀的,因此扰动可以写成如下形式:

    {{\boldsymbol{q}}'_{{\text{LST}}}}\left( {\xi ,\eta ,\zeta ,t} \right) = \tilde {\boldsymbol{q}}\left( \eta \right){{\mathrm{e}}^{{\text{i}}\left( {\alpha \xi + \beta \eta - \omega t} \right)}} + c.c. (2)

    其中:\alpha 是复数,\alpha = {\alpha _{\text{r}}} + {\text{i}}{\alpha _{\text{i}}}{\alpha _{\text{r}}}表示流向波数,{\alpha _{\mathrm{i}}}为增长率;\beta 为展向波数;\omega 为有量纲频率{f^*}的角频率,文中带有上标“*”的变量均为有量纲变量;\tilde {\boldsymbol{q}}为模态形状函数;c.c.表示共轭项。将上式代入Navier-Stokes方程,忽略非平行和非线性项,得到一维特征值方程,并采用求解特征值问题的Muller法进行求解。

    假设扰动从上游中性点位置(扰动开始放大处)[40]向下游演化,其扰动幅值可表示为:

    A\left( x \right) = {A_0}{{\text{e}}^{ - \displaystyle\int_{{x_0}}^x {{\alpha _{\mathrm{i}}}} {\text{d}}l}} (3)

    其中l为扰动传播方向。则N值表示为:

    N = \ln \left( {\dfrac{{{A_x}}}{{{A_0}}}} \right) = - \int_{{x_0}}^x {{\alpha _{\mathrm{i}}}} {\text{d}}l (4)

    本文使用的LST求解器是在国家数值风洞(National Numerical Wind Tunnel, NNW)项目支撑下联合天津大学开发的转捩预测代码[41]

    BiGlobal考虑了基本流的展向变化,小扰动可以写为:

    {{\boldsymbol{q}}'_{{\text{BiGlobal}}}}\left( {\xi ,\eta ,\zeta ,t} \right) = \tilde {\boldsymbol{q}}\left( {\eta ,\zeta } \right){{\mathrm{e}}^{{\text{i}}\left( {\alpha \xi - \omega t} \right)}} + c.c. (5)

    同样将式(1)和(5)代入Navier-Stokes方程,忽略非平行性和非线性项得到二维特征值方程:

    { \mathcal{\boldsymbol A}}\hat {\boldsymbol{q}} = \alpha \mathcal{\boldsymbol B}\hat {\boldsymbol{q}} (6)

    其中,\hat {\boldsymbol{q}} = [\tilde {\boldsymbol{q}},\alpha \tilde {\boldsymbol{q}}]{{\mathcal{\boldsymbol A}}}{{\mathcal{\boldsymbol B}}}为矩阵算子。

    计算域为半个锥的展向截面,迎风中心线为对称边界条件,壁面和上边界条件都为零边界条件。采用四阶中心差分对周向和法向进行离散,采用Arnoldi算法[42]获得特征值。

    PSE3D同时考虑了基本流展向变化及流向非平行性,可将小扰动写为:

    {{\boldsymbol{q}}'_{{\text{PSE3D}}}}\left( {\xi ,\eta ,\zeta ,t} \right) = \tilde {\boldsymbol{q}}\left( {\xi ,\eta ,\zeta } \right){{\text{e}}^{{\text{i}}\left( {\textstyle\int_\xi {\alpha \left( \xi \right){\text{d}}\xi - \omega t} } \right)}} + c.c. (7)

    将式(1)和(7)代入Navier-Stokes方程,忽略非线性项和\tilde {\boldsymbol{q}}\xi 的高阶导数项,得到面推进抛物化稳定性方程:

    {{\mathcal{L}}}\tilde {\boldsymbol{q}} + {{\mathcal{M}}}\frac{{\partial \tilde {\boldsymbol{q}}}}{{\partial \xi }} = 0 (8)

    其中 {{\mathcal{L}}} {{\mathcal{M}}} 都为线性算子,其完整展开形式参考文献[33]。\tilde {\boldsymbol{q}}\alpha 都与\xi 相关,在推进求解过程中,每个站位的\alpha 需要进行更新:

    \begin{split} {\alpha _{{\text{new}}}} = &- {\mathrm{i}}\frac{1}{E}\iint {{{\tilde \rho }^\dagger }}\frac{{\partial \tilde \rho }}{{\partial \xi }}\frac{{\bar T}}{{\bar \rho \gamma {Ma^2}}}{\left| {\tilde \rho } \right|^2} + \dfrac{{\bar \rho {{\tilde T}^\dagger }\dfrac{{\partial \tilde T}}{{\partial \xi }}}}{{\gamma \left( {\gamma - 1} \right){Ma^2}\bar T}} +\\ &\bar \rho \left( {{{\tilde u}^\dagger }\dfrac{{\partial \tilde u}}{{\partial \xi }} + {{\tilde v}^\dagger }\dfrac{{\partial \tilde v}}{{\partial \xi }} + {{\tilde w}^\dagger }\dfrac{{\partial \tilde w}}{{\partial \xi }}} \right){\text{d}}\eta {\text{d}}\zeta + {\alpha _{{\text{old}}}} \end{split} (9)

    其中,\rho T分别表示密度和温度,Ma为马赫数,\gamma 为比热比,上标“†”表示共轭,E为扰动能量,其表达式如式(10)所示:

    \begin{split} E = & \iint {\frac{{\bar T}}{{\bar \rho \gamma {Ma^2}}}}{\left| {\tilde \rho } \right|^2} + \bar \rho \left( {{{\left| {\tilde u} \right|}^2} + {{\left| {\tilde v} \right|}^2} + {{\left| {\tilde w} \right|}^2}} \right) +\\ & \frac{{{{\left| {\tilde T} \right|}^2}\bar \rho }}{{\gamma \left( {\gamma - 1} \right){Ma^2}\bar T}}{\text{d}}\eta {\text{d}}\zeta \end{split} (10)

    迭代更新\alpha 直至变化小于10–4,得到的模态增长率为:

    \sigma = - {\alpha _{\text{i}}} + \frac{1}{2}\frac{{{\text{d}}\ln E}}{{{\text{d}}\xi }} (11)

    N值可表示为:

    N\left( \xi \right) = \int_{{\xi _0}}^\xi \sigma {\text{d}}\xi (12)

    因为横流模态展向波长小,需要较高的网格分辨率,本文采用约化方法减少网格量[33]。仿照WKBJ(Wentzel-Kramers-Brillouin-Jeffery)的展开形式[43],将特征函数分解为慢变{\hat {\boldsymbol{q}}_m}和快变部分\exp \left( {i4{\text{π}} ms} \right),其中m为等效波数,s为从当地到背风中心线无量纲化的局部周长。特征方程变换为:

    {\mathcal{A}_m}{\hat {\boldsymbol{q}}_m} = \alpha {\mathcal{B}_m}{\hat {\boldsymbol{q}}_m} (13)

    当选取了合适的m后,新待求解的特征函数{\hat {\boldsymbol{q}}_m}沿展向慢变,分析所需的展向网格数将显著减少。约化方法的有效性和可靠性已在文献[32-33]中被验证,此处不再赘述。

    图2展示了m = 0、12和16时,x* = 250 mm处频率为35 kHz的横流模态的形状函数,其中Φ为周向角。可以看出,当m = 0时,展向变化剧烈,m = 12和16时,展向变化缓慢,其中m = 12展向变化最小,故后续讨论中选取m = 12。

    图  2  不同等效波数对应的模态形状函数(x* = 250 mm, f * = 35 kHz)
    Figure  2.  Modal shape functions of crossflow modes corresponding to different equivalent wavenumbers m (x* = 250 mm, f * = 35 kHz)

    图3(a)展示了x* = 128 mm处,BiGlobal分析得到的30 kHz非定常横流模态特征谱,其中相速度c定义为c = {\text{real}}\left( {{\omega \mathord{\left/ {\vphantom {\omega \alpha }} \right. } \alpha }} \right)\cos \left( {{7^ \circ }} \right)。可以看出,谱中包含多个不稳定模态,最不稳定模态(M5)增长率达19 /m。图3(b)展示了谱中M1和M5模态对应的无量纲流向速度形状函数,可以看出,模态主要分布在背风面100°~170°方位角之间,且模态间形状函数差别较小。

    图  3  特征谱和形状函数(x* = 128 mm,f * = 30 kHz,虚线为基本流\bar u \in [0∶0.15∶0.9])
    Figure  3.  Eigenvalue spectrum and modal shape functions of crossflow modes (x* = 128 mm, f* = 30 kHz, the baseflow is indicated by dashed lines of \bar u \in [0∶0.15∶0.9])

    为了对比不同模态向下游的演化情况,采用BiGlobal的结果作为PSE3D的入口条件,分析了30 kHz下不同相速度的非定常横流模态的演化特征,得到的N值演化如图4所示。可以看出,在x* = 200 mm的上游,各模态N值演化差别较小,随着x*的增加,相速度较小的模态N值逐渐超过相速度较大的模态。M1模态N值在模型尾端达到最大(接近7),因此后文对不同频率的模态分析都是在最小相速度模态下进行的。另外,为了验证PSE3D代码的可靠性,对M4模态的演化进行直接数值模拟,将M4模态的形状函数叠加到对应流向截面的基本流上,作为入口边界条件,其他边界条件与计算基本流时相同。在获得稳定的流场后,减去基本流,便可得到模态的演化。得到的模态N值演化如图4中黑色虚线所示。可以看出,在x* = 250~350 mm时,DNS和PSE3D得到的M4模态的N值增长趋势基本一致。

    图  4  30 kHz 下不同模态N值演化
    Figure  4.  Evolution of N factors of crossflow modes at 30 kHz

    图5(a)对比了PSE3D和DNS得到的30 kHz下 M4模态的温度扰动等值面,同时,还展示了LST得到的该频率下M4模态的N值等值线,其中T'为直接数值模拟得到的温度扰动量,T_{\rm r}为稳定性分析得到的温度形状函数的实部。可以看出,扰动为向背风中心线汇聚的条纹结构,向下游发展的过程中,扰动占据的空间向迎风面延拓,两种方法得到的等值面重合;M4模态的N值为8这一阵面在上游分布更宽,在下游与PSE3D得到的阵面基本一致。图5(b)对比了250 mm站位处周向角为120°和125°的温度剖面,可以看出,两种方法得到的温度剖面基本吻合,进一步验证了PSE3D代码的可靠性。

    图  5  30 kHz下,M4模态三维演化结构与LST得到的N值等值线对比
    Figure  5.  Comparison of three-dimensional evolution structure and the contour lines of N factors obtained by LST (f * = 30 kHz)

    图6展示了具有不同相速度模态的三维演化结构,可以看出,具有最小相速度的模态(M1)占据最宽的展向位置,随着相速度增大,模态在向下游演化的过程中,展向扩张趋势减弱,但波角逐渐增大。根据相速度的定义,大相速度对应着小流向波数,因此在初始站位具有较大相速度的M5模态沿下游演化过程中,波角始终大于其他模态。

    图  6  30 kHz下不同模态的三维演化结构
    Figure  6.  Three-dimensional evolution structures of different modes at 30 kHz

    图7展示了在x* = 128 mm处BiGlobal得到的两类模态的增长率和相速度随频率的变化规律:一类为对应频率下的最不稳定模态,一类为所在频率下的最小相速度模态。可以看出,随着模态频率从10 kHz增加到70 kHz,两类模态的相速度均从0.2左右逐渐增大到0.55附近,与风洞实验结果一致[7]。两类模态最不稳定频率均在30 kHz附近。相比而言,最小相速度模态的增长率较小,可近似视为中性稳定模态。

    图  7  x* = 128 mm处不同频率模态的增长率和相速度分布
    Figure  7.  Growth rates and phase velocities of modes with different frequencies at x* = 128 mm

    x* = 128 mm处不同频率的最小相速度模态作为初值代入PSE3D中计算不同频率的非定常横流模态的N值演化,如图8所示。可以看出,30 kHz的非定常横流模态幅值增长最快,其N值最大可达7左右,高频横流模态(50 ~70 kHz)较为稳定。图中黑线对应LST获得的N值包络,可以看到,LST得到的N值远大于PSE3D得到的模态N值。造成这种差异的原因可能有如下3点:(1)积分路径不同。LST沿势流方向积分,而PSE3D则沿流向积分,同时由于LST积分起点更靠近上游,故LST积分路径更长;(2)两者增长率定义不同。LST的增长率为当地最不稳定模态增长率,衡量的是积分路径两点之间的扰动能量变化率,而PSE3D则考虑两个相邻流向截面之间的扰动能量变化率;(3)流场假设不同。LST假设流动在展向上是均匀的,因此不同展向位置处的扰动为独立演化,而PSE3D考虑了流场的展向变化,其扰动在展向以耦合的波包形式出现。

    图  8  不同频率非定常横流模态N值演化与LST获得的N值包络对比
    Figure  8.  Evolution of N factors and comparison of the N factor envelope solved by LST of modes with different frequencies

    图9(a、b)分别展示了20 kHz和40 kHz横流模态的三维演化结构以及对应频率通过LST得到的N值等值线。可以看出,两个频率模态的波角相近,但低频模态更靠近背风中心线,其展向波长更大;N = 8的等值线在上游占据更宽的展向空间,但在下游,展向宽度与PSE3D的结果吻合较好。

    图  9  横流模态(20 kHz和40 kHz)三维演化结构与LST得到的N值等值线对比
    Figure  9.  Comparison of three-dimensional evolution structures of crossflow modes (20 kHz and 40 kHz) with contour lines of N factors obtained by LST

    图10分别展示了x* = 200 、250、300、350 mm站位处,PSE3D和BiGloba1得到的非定常横流模态N值对比,两种方法得到的N值积分起点均为x* = 128 mm站位处。其中,BiGlobal方法的N值由当地最不稳定模态增长率积分得到。从图中可以看出,在x* = 200 mm处,两种方法得到的N值都较小,但得到的主频存在差异。在下游x* = 250 mm和300 mm处,两种方法得到的N值接近,且主频一致,这是由于最小相速度模态M1在下游x* = 200 mm处逐渐转变为最不稳定模态(如图4所示)。在x* = 350 mm处,虽主频一致,但两种方法得到的N值差异增大,造成这种差异的原因可能是PSE3D计算的是x* = 128 mm处相速度最小的不稳定模态,且考虑了流向非平行性,而BiGlobal的结果对应最不稳定模态,两个模态没有一一对应。

    图  10  不同流向站位处BiGlobal和PSE3D的N值对比
    Figure  10.  Comparison of N factors obtained by BiGlobal and PSE3Dat different locations

    图11对比了PSE3D、BiGlobal和LST得到的准定常横流模态(f * = 0.1 kHz)的N值。可以看出,PSE3D得到的准定常横流模态N值最大可达4.7,低于最不稳定非定常横流模态。BiGlobal和PSE3D的N值较接近,在x* = 200 mm上游,两者几乎重合,随着向下游发展,两种方法得到的N值差别逐渐增大。LST得到的N值显著大于PSE3D和BiGlobal得到的N值。图12展示了准定常横流模态的演化三维结构,可以看出,相比非定常横流模态,准定常横流模态波角更大。

    图  11  准定常横流模态N值演化
    Figure  11.  N factor evolution of quasi-stationary crossflow modes
    图  12  准定常横流模态三维演化结构与LST得到的N值等值线对比
    Figure  12.  Comparison of three-dimensional evolution structure of the quasi-stationary crossflow mode and comparison of the contour lines of N factors obtained by LST

    本文采用BiGlobal和PSE3D两种全局稳定性分析方法对马赫数6常规风洞实验工况下的6°攻角圆锥横流失稳区进行了分析,并与LST和DNS结果进行了对比,得到如下结论:

    1) 70 kHz以下的横流模态更不稳定,其中准定常横流模态较20~40 kHz的非定常横流模态稳定。横流模态主要分布在背风面,在下游逐渐向迎风面扩展。准定常横流模态的波角较非定常横流模态大。

    2) DNS验证了PSE3D得到的模态幅值和形状函数的演化特征。LST得到的N值总是大于BiGlobal和PSE3D两种分析方法的结果,但得到的N值转捩阵面与PSE3D得到的模态形状函数等值面接近。

    本文基于全局稳定性分析方法得到了横流模态的三维演化结构以及幅值演化特征,对扰动线性演化有了更加直观的认识。接下来将进一步对横流二次失稳和强非线性过程进行研究,深入理解横流失稳机制。

  • 图  1   计算域和基本流

    Figure  1.   Computational domain and baseflow

    图  2   不同等效波数对应的模态形状函数(x* = 250 mm, f * = 35 kHz)

    Figure  2.   Modal shape functions of crossflow modes corresponding to different equivalent wavenumbers m (x* = 250 mm, f * = 35 kHz)

    图  3   特征谱和形状函数(x* = 128 mm,f * = 30 kHz,虚线为基本流\bar u \in [0∶0.15∶0.9])

    Figure  3.   Eigenvalue spectrum and modal shape functions of crossflow modes (x* = 128 mm, f* = 30 kHz, the baseflow is indicated by dashed lines of \bar u \in [0∶0.15∶0.9])

    图  4   30 kHz 下不同模态N值演化

    Figure  4.   Evolution of N factors of crossflow modes at 30 kHz

    图  5   30 kHz下,M4模态三维演化结构与LST得到的N值等值线对比

    Figure  5.   Comparison of three-dimensional evolution structure and the contour lines of N factors obtained by LST (f * = 30 kHz)

    图  6   30 kHz下不同模态的三维演化结构

    Figure  6.   Three-dimensional evolution structures of different modes at 30 kHz

    图  7   x* = 128 mm处不同频率模态的增长率和相速度分布

    Figure  7.   Growth rates and phase velocities of modes with different frequencies at x* = 128 mm

    图  8   不同频率非定常横流模态N值演化与LST获得的N值包络对比

    Figure  8.   Evolution of N factors and comparison of the N factor envelope solved by LST of modes with different frequencies

    图  9   横流模态(20 kHz和40 kHz)三维演化结构与LST得到的N值等值线对比

    Figure  9.   Comparison of three-dimensional evolution structures of crossflow modes (20 kHz and 40 kHz) with contour lines of N factors obtained by LST

    图  10   不同流向站位处BiGlobal和PSE3D的N值对比

    Figure  10.   Comparison of N factors obtained by BiGlobal and PSE3Dat different locations

    图  11   准定常横流模态N值演化

    Figure  11.   N factor evolution of quasi-stationary crossflow modes

    图  12   准定常横流模态三维演化结构与LST得到的N值等值线对比

    Figure  12.   Comparison of three-dimensional evolution structure of the quasi-stationary crossflow mode and comparison of the contour lines of N factors obtained by LST

  • [1] 陈坚强, 涂国华, 张毅锋, 等. 高超声速边界层转捩研究现状与发展趋势[J]. 空气动力学学报, 2015, 35(3): 311337. doi: 10.7638/kqdlxxb-2017.0030

    CHEN J Q, TU G H, ZHANG Y F, et al. Hypersonic boundary layer transition: What we know, where shall we go[J]. Acta Aerodynamica Sinica, 2015, 35(3): 311337(in Chinese). doi: 10.7638/kqdlxxb-2017.0030

    [2] 苏彩虹. 高超声速边界层转捩预测中的关键科学问题: 感受性、扰动演化及转捩判据研究进展[J]. 空气动力学学报, 2020, 38(2): 355367. doi: 10.7638/kqdlxxb-2020.0056

    SU C H. Progress in key scientific problems of hypersonic bounary-layer transition prediction: Receptivity, evolution of disturbances and transition criterion[J]. Acta Aerodynamica Sinica, 2020, 38(2): 355367(in Chinese). doi: 10.7638/kqdlxxb-2020.0056

    [3]

    LEE C B, CHEN S Y. Recent progress in the study of transition in the hypersonic boundary layer[J]. National Science Review, 2019, 6(1): 155170. doi: 10.1093/nsr/nwy052

    [4]

    HENDERSON R O. Crossflow transition at Mach 6 on a cone at low angles of attack[D]. USA: Purdue University, 2014.

    [5]

    NEEL I T, LEIDY A, TICHENOR N R, et al. Influence of environmental disturbances on hypersonic crossflow instability on the HIFiRE-5 elliptic cone[C]// 2018 AIAA Aerospace Sciences Meeting, Kissimmee, Florida. Reston, Virginia: AIAA, 2018.

    [6]

    HEMBLING E, WIRTH J, SEMPER M. Hypersonic crossflow investigation on the HIFiRE-5 elliptic cone[C]// AIAA Scitech 2021 Forum, VIRTUAL EVENT. Reston, Virginia: AIAA, 2021.

    [7]

    WARD C A C. Crossflow instability and transition on a circular cone at angle of attack in a Mach-6 quiet tunnel[D]. USA: Purdue University, 2014.

    [8]

    WARD C, HENDERSON R, SCHNEIDER S P. Secondary instability of stationary crossflow vortices on an inclined cone at Mach 6[C]// 45th AIAA Fluid Dynamics Conference, Dallas, TX. Reston, Virginia: AIAA, 2015.

    [9]

    XU G L, LIU G, CHEN J Q, et al. Role of freestream slow acoustic waves in a hypersonic three-dimensional boundary layer[J]. AIAA Journal, 2018, 56(9): 35703584. doi: 10.2514/1.J056492

    [10]

    ADAMCZAK D, KIMMEL R, PAULL A, et al. HIFiRE-1 flight trajectory estimation and initial experimental results[C]//17th AIAA International Space Planes and Hypersonic Systems and Technologies Conference, San Francisco, California. Reston, Virginia: AIAA, 2011.

    [11]

    KIMMEL R, ADAMCZAK D, PAULL A, et al. HIFiRE-1 preliminary aerothermodynamic measurements[C]// 41st AIAA Fluid Dynamics Conference and Exhibit, Honolulu, Hawaii. Reston, Virginia: AIAA, 2011.

    [12]

    WAN B B, TU G H, YUAN X X, et al. Identification of traveling crossflow waves under real hypersonic flight conditions[J]. Physics of Fluids, 2021, 33(4): 044110. doi: 10.1063/5.0046954

    [13]

    KIMMEL R L, ADAMCZAK D W, BORG M P, et al. First and fifth hypersonic international flight research experimentation's flight and ground tests[J]. Journal of Spacecraft and Rockets, 2019, 56(2): 421431. doi: 10.2514/1.a34287

    [14] 陈久芬, 凌岗, 张庆虎, 等. 7°尖锥高超声速边界层转捩红外测量实验[J]. 实验流体力学, 2020, 34(1): 6066.

    CHEN J F, LING G, ZHANG Q H, et al. Infrared thermography experiments of hypersonic boundary-layer transition on a 7° half-angle sharp cone[J]. Journal of Experiments in Fluid Mechanics, 2020, 34(1): 6066(in Chinese).

    [15]

    YATES H B, TUFTS M W, JULIANO T J. Analysis of the hypersonic cross-flow instability with experimental wavenumber distributions[J]. Journal of Fluid Mechanics, 2020, 883: A50. doi: 10.1017/jfm.2019.864

    [16]

    YATES H B, MATLIS E H, JULIANO T J, et al. Plasma-actuated flow control of hypersonic crossflow-induced boundary-layer transition[J]. AIAA Journal, 2020, 58(5): 20932108. doi: 10.2514/1.j058981

    [17]

    CRAIG S A, SARIC W S. Crossflow instability in a hypersonic boundary layer[J]. Journal of Fluid Mechanics, 2016, 808: 224244. doi: 10.1017/jfm.2016.643

    [18]

    CRAIG S A, SARIC W S. Crossflow instability on a yawed cone at Mach 6[J]. Procedia IUTAM, 2015, 14: 1525. doi: 10.1016/j.piutam.2015.03.019

    [19]

    DONG S W, CHEN J Q, YUAN X X, et al. Wall pressure beneath a transitional hypersonic boundary layer over an inclined straight circular cone[J]. Advances in Aerodynamics, 2020, 2(1): 29. doi: 10.1186/s42774-020-00057-4

    [20]

    WANG Q, XIANG X H, DONG S W, et al. Wall temperature effects on the hypersonic boundary-layer transition over an inclined, blunt cone[J]. Physics of Fluids, 2023, 35(2): 024107. doi: 10.1063/5.0136754

    [21]

    SARIC W S, REED H L, WHITE E B. Stability and transition of three-dimensional boundary layers[J]. Annual Review of Fluid Mechanics, 2003, 35: 413440. doi: 10.1146/annurev.fluid.35.101101.161045

    [22]

    PEREZ E, REED H L, KUEHL J. Instabilities on a hypersonic yawed straight cone[C]// 43rd Fluid Dynamics Conference, San Diego, CA. Reston, Virginia: AIAA, 2013.

    [23] 苏彩虹, 周恒. 超音速和高超音速有攻角圆锥边界层的转捩预测[J]. 中国科学G辑, 2009, 39(6): 874–882.

    SU C H, ZHOU H. Transition prediction of supersonic and hypersonic cone boundary layer with angle of attack[J]. Scientia Sinica (Physica, Mechanica & Astronomica), 2009, 39(6): 874–882(in Chinese).

    [24] 袁湘江, 李国良, 刘智勇, 等. 小攻角高超声速钝锥边界层失稳特性[J]. 航空动力学报, 2011, 26(12): 28052811.

    YUAN X J, LI G L, LIU Z Y, et al. Study of the instability characteristic in the boundary layer of a hypersonic blunt cone at low angle of attack[J]. Journal of Aerospace Power, 2011, 26(12): 28052811(in Chinese).

    [25] 宋润杰. 高超声速三维边界层的流动稳定性及扰动演化[D]. 天津: 天津大学, 2020.

    SONG R J. Flow stability and disturbance evolution of hypersonic three-dimensional boundary layer[D]. Tianjin: Tianjin University, 2020(in Chinese).

    [26]

    SONG R J, ZHAO L, HUANG Z F. Improvement of the parabolized stability equation to predict the linear evolution of disturbances in three-dimensional boundary layers based on ray tracing theory[J]. Physical Review Fluids, 2020, 5(3): 033901. doi: 10.1103/physrevfluids.5.033901

    [27] 黄章峰, 万兵兵, 段茂昌. 高超声速流动稳定性及转捩工程应用若干研究进展[J]. 空气动力学学报, 2020, 38(2): 368378. doi: 10.7638/kqdlxxb-2020.0047

    HUANG Z F, WAN B B, DUAN M C. Progresses in engineering application research on hypersonic flow stability and transition[J]. Acta Aerodynamica Sinica, 2020, 38(2): 368378(in Chinese). doi: 10.7638/kqdlxxb-2020.0047

    [28]

    THEOFILIS V. Global linear instability[J]. Annual Review of Fluid Mechanics, 2011, 43: 319352. doi: 10.1146/annurev-fluid-122109-160705

    [29]

    DE TULLIO N, PAREDES P, SANDHAM N D, et al. Laminar–turbulent transition induced by a discrete roughness element in a supersonic boundary layer[J]. Journal of Fluid Mechanics, 2013, 735: 613646. doi: 10.1017/jfm.2013.520

    [30]

    PAREDES P, GOSSE R, THEOFILIS V, et al. Linear modal instabilities of hypersonic flow over an elliptic cone[J]. Journal of Fluid Mechanics, 2016, 804: 442466. doi: 10.1017/jfm.2016.536

    [31]

    LAKEBRINK M T, BORG M P. Traveling crossflow wave predictions on the HIFiRE-5 at Mach 6: Stability analysis vs. quiet tunnel data[C]// 54th AIAA Aerospace Sciences Meeting, San Diego, California, USA. Reston, Virginia: AIAA, 2016.

    [32]

    LIU S Y, WAN B B, YUAN X X, et al. Linear modal global instabilities of hypersonic flow over an inclined cone[J]. Physics of Fluids, 2022, 34(7): 074108. doi: 10.1063/5.0097358

    [33]

    CHEN X, DONG S W, TU G H, et al. Boundary layer transition and linear modal instabilities of hypersonic flow over a lifting body[J]. Journal of Fluid Mechanics, 2022, 938: A8. doi: 10.1017/jfm.2021.1125

    [34]

    CHEN X, DONG S W, TU G H, et al. Global stability analyses of Mack mode on the windward face of a hypersonic yawed cone[J]. Physical Review Fluids, 2023, 8(3): 033903. doi: 10.1103/PhysRevFluids.8.033903

    [35]

    ZHANG L, DONG S, LIU S, et al. Stability analysis of streamwise vortices over a blunt inclined cone under a hypersonic flight condition[J]. Physics of Fluids, 2022, 34(7): 074107. doi: 10.1063/5.0099952

    [36] 陈曦, 陈坚强, 董思卫, 等. 高超声速6°迎角圆锥边界层背风流向涡稳定性分析[J]. 空气动力学学报, 2020, 38(2): 299307. doi: 10.7638/kqdlxxb-2020.0038

    CHEN X, CHEN J Q, DONG S W, et al. Stability analyses of leeward streamwise vortices for a hypersonic yawed cone at 6 degree angle of attack[J]. Acta Aerodynamica Sinica, 2020, 38(2): 299307(in Chinese). doi: 10.7638/kqdlxxb-2020.0038

    [37] 李新亮. 高超声速湍流直接数值模拟技术[J]. 航空学报, 2015, 36(1): 147158.

    LI X L. Direct numerical simulation techniques for hypersonic turbulent flows[J]. Acta Aeronautica et Astronautica Sinica, 2015, 36(1): 147158(in Chinese).

    [38]

    LI X H, CHEN J Q, HUANG Z F, et al. Stability analysis and transition prediction of streamwise vortices over a yawed cone at Mach 6[J]. Physics of Fluids, 2020, 32(12): 124110. doi: 10.1063/5.0031057

    [39]

    ZHANG L G, WAN B B, DONG S W, et al. Wall temperature effects on the stability of leeward vortices of a blunt inclined cone[J]. Acta Mechanica Sinica, 2024, 40(7): 123563. doi: 10.1007/s10409-023-23563-x

    [40] 周恒, 赵耕夫. 流动稳定性[M]. 北京: 国防工业出版社, 2004: 66–67.

    ZHOU H, ZHAO G F. Hydrodynamic stability[M]. Beijing: National Defense Industry Press, 2004: 66–67(in Chinese).

    [41] 万兵兵, 陈曦, 陈坚强, 等. 三维边界层转捩预测HyTEN 软件在高超声速典型标模中的应用[J]. 空天技术, 2023(1): 150158. doi: 10.16338/j.issn.2097-0714.20220206

    WAN B B, CHEN X, CHEN J Q, et al. Applications of HyTEN software for predicting three-dimensional boundary-layer transition in typical hypersonic models[J]. Aerospace Technology, 2023(1): 150158(in Chinese). doi: 10.16338/j.issn.2097-0714.20220206

    [42]

    ARNOLDI W E. The principle of minimized iterations in the solution of the matrix eigenvalue problem[J]. Quarterly of Applied Mathematics, 1951, 9(1): 1729. doi: 10.1090/qam/42792

    [43]

    BENDER C M, ORSZAG S A. Advanced mathematical methods for scientists and engineers: Asymptotic methods and perturbation theory[M]. German: Springer, 1999.

图(12)
计量
  • 文章访问数:  28
  • HTML全文浏览量:  3
  • PDF下载量:  16
  • 被引次数: 0
出版历程
  • 收稿日期:  2024-06-25
  • 修回日期:  2024-08-07
  • 录用日期:  2024-08-19
  • 网络出版日期:  2024-10-29
  • 发布日期:  2024-10-29
  • 刊出日期:  2025-02-24

目录

/

返回文章
返回