Utafiti huu umeendeleza mbinu ya ulinganishaji wa mawimbi madogo ya Haar kwa ajili ya kukokotoa mtiririko wa hipasoniki wenye mabadiliko makali sana kama mawimbi ya mshtuko, tabaka za shear na mwingiliano wa wimbi la mshtuko/tabaka la mpaka. Mbinu hiyo imetumika kwenye milinganyo ya Navier–Stokes inayobanika ya vipimo vitatu, na ujumuishaji wa muda wa Runge–Kutta wa daraja la tatu umetumiwa. Watafiti waliijaribu mbinu hiyo kwa vortex ya isentropiki inayosafirishwa kwa muda mrefu, tabaka la mchanganyiko la Mach 7,5, mpito wa tabaka la mpaka wa vipimo vitatu wa Mach 6 na kona ya mgandamizo ya Mach 6. Matokeo yanaonyesha kwamba mbinu inaweza kuwakilisha mabadiliko makali bila kuunda mitikisiko bandia ya aina ya Gibbs karibu na kutokuwa endelevu, haiyafifishi machafuko ya amplitudo ndogo kwa damping ya namba iliyofichika, na inaweza kuzalisha upya uundaji wa vortex, mpito wa mtiririko na gradients za joto karibu na ukuta katika mifano teule ya hipasoniki. Hata hivyo, tathmini zinategemea kwa kiasi kikubwa kulinganisha kwa ubora miundo ya mtiririko; hakuna kipimo cha majaribio, error norm ya kina wala ulinganisho wa namba wa gharama sawa dhidi ya mbinu za kisasa za kunasa mshtuko.

Faida kuu ya mawimbi madogo ya Haar ni kuwa na msingi wa vipande usiobadilika unaoendana kiasili na functions zenye umbo la hatua au zinazobadilika kwa kasi. Global spectral bases za kawaida zinaweza kuunda mitikisiko ya high-frequency karibu na kutokuwa endelevu ambayo haipo katika suluhisho halisi. Haar basis, kwa upande mwingine, iliwasilisha derivatives ya kwanza na ya pili bila Gibbs oscillations katika mfano wa Heaviside step function uliotumika katika utafiti. Mbinu haiongezi implicit numerical diffusion katika maeneo yasiyo na strong gradients; katika maeneo yanayohitaji uthabiti, kama shock interactions, hutumia explicit artificial viscosity inayopungua kadiri grid inavyoboreshwa.

Tathmini kwa mtazamo wa Uturuki: Mbinu inaweza kubadilishwa kwa miundombinu ya computational fluid dynamics inayotengenezwa nchini Uturuki kwa ajili ya hypersonic air vehicles, re-entry systems, high-speed aerothermodynamics, boundary-layer transition, shock-wave/surface interactions na thermal-protection-system research. Ili itumike katika universities, research centers na aerospace organizations, algorithm inapaswa kutekelezwa katika local solvers, kuzalishwa upya kwa open reference tests, kuthibitishwa dhidi ya wind-tunnel au published experimental data, parallel-computing performance kupimwa, na kulinganishwa na WENO, TENO, finite-volume au discontinuous Galerkin methods katika grid na error level zilezile. Kwa kuwa utafiti haukuchunguza vehicle geometry iliyotengenezwa Uturuki, local flight test, material system au operational design, matokeo mahususi ya Uturuki kuhusu performance, safety au cost hayawezi kutolewa moja kwa moja.

Kwa nini suluhisho la namba la mtiririko wa hipasoniki ni gumu?

Katika hypersonic flow fields, spatial scales tofauti sana zinaweza kuwepo kwa wakati mmoja. Wakati overall length ya vehicle inaweza kuelezwa kwa meters, mabadiliko katika shock layer yanaweza kutokea katika micrometer scale. Utafiti unasema kwamba typical shock width inaweza kuwa takribani katika kiwango cha 7–9 micrometers na thermodynamic variables zinaweza kuonyesha sudden jump ndani ya eneo hili nyembamba sana.

Miundo ifuatayo inaweza kuwepo pamoja katika flow field ileile:

• Shock waves,
• Compression na expansion regions,
• Shear layers,
• Shock-shock interactions,
• Shock-wave/boundary-layer interactions,
• Small-amplitude instability waves,
• Laminar-to-turbulent transition structures,
• Strong temperature na density gradients karibu na ukuta.

Ili numerical method iweze kutatua mshtuko kwa uthabiti, lazima ikandamize artificial oscillations karibu na sharp changes. Hata hivyo, numerical diffusion nyingi inaweza pia kuzima very-small-amplitude waves zinazoanzisha transition. Kwa hiyo, katika hypersonic-transition calculations mahitaji mawili yanashindana: stability karibu na shock na low damping katika smooth regions.

Changamoto kuu za mbinu zilizopo

ENO, WENO na TENO methods katika finite-difference family hutumika sana katika flows zenye shocks. Methods hizi hubadilisha computational stencil au weights kulingana na smoothness ya solution ili kupunguza oscillations karibu na discontinuity. Compact finite-difference methods zinaweza kutoa near-spectral resolution katika smooth regions.

Finite-element na discontinuous Galerkin methods zinaweza kutumika kwa urahisi zaidi kwenye complex geometries na irregular meshes. Hata hivyo, katika high-order elements, sensor-dependent artificial viscosity au stabilization mechanisms nyingine zinahitajika kwa shock capturing. Pia, degrees of freedom nyingi zinaweza kuhitajika katika kila element ili kutatua transition waves zenye high Mach numbers.

Classical spectral methods zinaweza kutoa high accuracy kwa degrees of freedom chache katika smooth functions. Lakini global trigonometric au polynomial basis huunda artificial oscillations karibu na discontinuity kutokana na Gibbs phenomenon. Strong filtering au artificial diffusion inaweza kupunguza oscillations hizi, lakini inaweza pia kusababisha small physical instabilities kupotea.

Mbinu ya wavelet inabadilisha nini?

Wavelet methods hutumia forms za mother wavelet zilizohamishwa kwenye scales na locations tofauti badala ya kudhani global smoothness moja. Kwa njia hii, spatial location na scale information vinaweza kuhifadhiwa. Ikiwa local structure katika solution inafanana na wavelet basis inayotumika, inaweza kuwakilishwa kwa coefficients chache.

Haar wavelet ndiyo familia rahisi zaidi ya wavelet. Umbo lake la msingi ni +1 katika interval fulani, −1 katika interval iliyo karibu na zero katika maeneo mengine. Muundo huu wa piecewise constant na discontinuous unafaa kiasili kwa sharp transitions zinazofanana na shock.

Michango asili ya utafiti imewasilishwa chini ya vichwa vitatu:

1. Kutengeneza mathematical formulation kamili ya Haar wavelet collocation method kwa three-dimensional compressible Navier–Stokes equations,
2. Kuonyesha kwa long-time vortex test kwamba mbinu haitoi implicit numerical damping,
3. Kuitumia kwa hypersonic flows kama Mach 7,5 shear layer, Mach 6 boundary-layer transition na Mach 6 compression corner.

Gibbs phenomenon ni nini?

Gibbs phenomenon ni artificial oscillations zinazotokea karibu na jump point wakati discontinuous function inawakilishwa kwa makadirio kwa smooth na global bases. Grid au idadi ya terms ikiongezeka, eneo la oscillations linaweza kuwa nyembamba; lakini peak amplitude inaweza kutotoweka kabisa.

Katika utafiti, Heaviside step function imetumika kuwakilisha thermodynamic variable inayobadilika ghafla kupitia shock:

\[ H_s(x)= \begin{cases} 1, & x\geq 0{,}5,\\ 0, & x<0{,}5. \end{cases} \]

Second derivative ya function katika Haar wavelet basis imeelezwa kama:

\[ \frac{d^2H_s}{dx^2}=\sum_{j=1}^{2M}a_jh_j(x) \]

Hapa \(h_j(x)\) inaonyesha Haar basis functions, na \(a_j\) wavelet coefficients.

Baada ya ku-integrate mara mbili:

\[ H_s=Q\mathbf{a}+Ax+B \]

expression hupatikana. Boundary conditions zinapotumika, coefficient vector hukokotolewa kama:

\[ \mathbf{a}=\left[Q-xQ\big|_{x=1}\right]^{-1}(H_s-x) \]

Hukokotolewa kwa njia hii.

Kielelezo 1 kinalinganisha Haar method na Chebyshev collocation method. Chebyshev solution huzalisha artificial oscillations kubwa sana pande zote mbili za discontinuity na hasa katika second derivative, wakati Haar method haionyeshi oscillations hizi. Hii ndiyo mathematical advantage kuu ambayo utafiti unategemea kwa hypersonic flows.

Haar wavelet basis imefafanuliwaje?

Katika interval \(x\in[0,1]\), Haar wavelet ni:

\[ h_i(x)= \begin{cases} 1, & x\in[\xi_1,\xi_2),\\ -1, & x\in[\xi_2,\xi_3),\\ 0, & \text{diğer durumlarda} \end{cases} \]

Subinterval boundaries zimetolewa kama:

\[ \xi_1=\frac{k}{m},\qquad \xi_2=\frac{k+0{,}5}{m},\qquad \xi_3=\frac{k+1}{m} \]

\(m=2^j\) huamua expansion level, na \(k\) huamua location ya wavelet katika space. Kadiri resolution level \(J\) inavyoongezeka, Haar functions za smaller scales huongezwa na total number ya represented wavelets huongezeka.

Kielelezo B.8 kinaonyesha Haar wavelets tatu za kwanza. Function ya kwanza ni constant katika interval yote, huku functions zinazofuata zikichukua values za +1 na −1 katika intervals ndogo zaidi. Hierarchical structure hii inaruhusu large-scale na small-scale features kuwakilishwa katika basis ileile.

Kwa nini derivative ya juu zaidi imewakilishwa moja kwa moja kwa wavelets?

Kwa kuwa Haar functions ni discontinuous, kuchukua derivatives zake kwa classical sense si sahihi. Katika utafiti, tatizo hili limetatuliwa kwa mwelekeo wa kinyume: highest-order derivative katika differential equation imeandikwa katika Haar basis, kisha expression hii imebadilishwa kuwa original function kupitia integration.

Integral ya Haar function mara \(r\) imefafanuliwa kama:

\[ h_{r,i}(x)=\int_0^x\int_0^x\cdots\int_0^x h_i(t)\,dt^r \]

Kwa hiyo, badala ya kuchukua derivative moja kwa moja ya discontinuous basis, analytically defined integration matrices zimetumika.

Three-dimensional Navier–Stokes equations

Mbinu imetumika kwenye three-dimensional compressible Navier–Stokes equations kwa calorically perfect gas:

\[ \frac{\partial\mathbf{w}}{\partial t} +\frac{\partial\mathbf{F}_i}{\partial x} +\frac{\partial\mathbf{G}_i}{\partial y} +\frac{\partial\mathbf{H}_i}{\partial z} = \frac{\partial\mathbf{F}_v}{\partial x} +\frac{\partial\mathbf{G}_v}{\partial y} +\frac{\partial\mathbf{H}_v}{\partial z}. \]

Vector ya conserved variables imetolewa kama:

\[ \mathbf{w}=(\rho,\rho u,\rho v,\rho w,E)^T \]

Hapa \(\rho\) ni density, \(u\), \(v\) na \(w\) ni Cartesian velocity components, na \(E\) ni total energy density.

Pressure hupatikana kwa:

\[ p=(\gamma-1)\left[E-\frac{1}{2}\rho(u^2+v^2+w^2)\right] \]

Hupatikana kwa uhusiano huu.

Viscous stress tensor imefafanuliwa kama:

\[ \sigma_{ij}=\mu\left( \frac{\partial u_i}{\partial x_j} +\frac{\partial u_j}{\partial x_i} \right) -\frac{2}{3}\mu\delta_{ij} \frac{\partial u_k}{\partial x_k} \]

na heat flux kama:

\[ q_i=-k\frac{\partial T}{\partial x_i} \]

Imefafanuliwa kwa namna hii.

Thermal conductivity na temperature zimekokotolewa kwa relations:

\[ k=\frac{C_p\mu}{Pr},\qquad T=\frac{p}{R\rho} \]

Kwa viscosity, temperature-dependent power law ilitumika:

\[ \frac{\mu}{\mu_{ref}}= \left(\frac{T}{T_{ref}}\right)^\omega \]

na source inatoa values \(\mu_{ref}=1{,}716\times10^{-5}\), \(T_{ref}=273\) na \(\omega=0{,}7\). Unit ya viscosity haijaandikwa tofauti katika source equation.

Spatial discretization ilifanywaje?

Calculation ilifanywa katika coordinates \((\xi,\eta,\zeta)\), na independent resolution levels \(I\), \(J\) na \(K\) zilitumika kwa directions tatu. Katika kila time step, Haar coefficients za conserved variables hukokotolewa tofauti katika directions tatu.

Kwa mfano, coefficient vector katika direction ya \(\xi\) imetolewa kama:

\[ \mathbf{a}_{\xi} =MQ_{\xi}^{-1} \left[ \mathbf{w}-\mathbf{w}_L -\xi(\mathbf{w}_R-\mathbf{w}_L) \right] \]

Hapa \(\mathbf{w}_L\) na \(\mathbf{w}_R\) ni left na right boundary values katika direction husika.

First derivative hukokotolewa kama:

\[ \frac{\partial\mathbf{w}}{\partial\xi} =MP_\xi\mathbf{a}_\xi +(\mathbf{w}_R-\mathbf{w}_L) \]

na second derivative kama:

\[ \frac{\partial^2\mathbf{w}}{\partial\xi^2} =MH_\xi\mathbf{a}_\xi \]

Derivatives katika computational coordinates kisha hubadilishwa kuwa physical \(x\), \(y\) na \(z\) derivatives kupitia grid transformation iliyotumika.

Time integration

Utafiti ulitumia third-order Runge–Kutta method:

\[ \mathbf{w}_1= \Delta t\frac{\partial\mathbf{w}_0}{\partial t} +\mathbf{w}_0 \]

\[ \mathbf{w}_2= \frac{1}{4} \left( \Delta t\frac{\partial\mathbf{w}_1}{\partial t} +\mathbf{w}_1 \right) +\frac{3}{4}\mathbf{w}_0 \]

\[ \mathbf{w}_3= \frac{2}{3} \left( \Delta t\frac{\partial\mathbf{w}_2}{\partial t} +\mathbf{w}_2 \right) +\frac{1}{3}\mathbf{w}_0. \]

\(\mathbf{w}_0\) inawakilisha solution katika previous time step. Katika kila validation case, time step iliamuliwa kulingana na CFL number iliyotajwa.

Je, mbinu haitumii artificial diffusion kabisa?

Katika utafiti, dhana mbili tofauti zinapaswa kutenganishwa:

• Inadaiwa kwamba hakuna implicit numerical diffusion inayotokana na structure yenyewe ya Haar discretization.
• Katika calculations zenye shock au strong gradient, explicit artificial viscosity huongezwa kwa stability.

Coefficient ya explicit artificial viscosity ilichaguliwa kuwa inversely proportional na grid resolution. Kwa hiyo, artificial viscosity inalengwa kupungua kadiri grid inavyoboreshwa. Hata hivyo, source haitoi detailed list ya coefficients zilizotumika kwa tests zote, sensor form wala sensitivity analysis ya solution quality. Kwa hiyo, kauli “hakuna numerical diffusion kabisa” inahusu tu implicit behavior ya method; explicit stabilization term inatumika katika strong-gradient tests.

Umuhimu wa H, P na Q matrices

Katika Haar approach, \(H\) inawakilisha basis values, \(P\) first integrals na \(Q\) second integrals. Kwa kuwa highest spatial derivative katika Navier–Stokes equations ni second order, \(Q\) matrix ndiyo minimum integration level inayohitajika.

Kielelezo 2 na Kielelezo 3 vinaonyesha nonzero elements za matrices hizi na inverses zake. Forward \(H\) na \(P\) matrices ni sparse sana. \(Q\) na hasa inverse matrices zinazotumika katika coefficient calculation ni dense zaidi. \(H\) matrix ni orthogonal; baada ya integration, property hii inapotea.

Implementation ya sasa ilitumia dense matrix-vector multiplication kwa lengo la kuonyesha accuracy. Computational cost ya implementation hii katika kila spatial direction ni takribani:

\[ O(N_i^2) \]

kwa scale hiyo na parallel efficiency ni limited. Watafiti wanasema fast wavelet transforms zinaweza kupunguza cost hii kinadharia hadi:

\[ O(N_i) \]

order. Hata hivyo, fast algorithm haikutekelezwa wala kupimwa katika utafiti huu; imewasilishwa kama direction ya future development.

Jaribio la kwanza: Vortex ya isentropiki ya muda mrefu

Jaribio la kwanza lilichunguza kama vortex ya isentropiki isiyosogea inaweza kuhifadhiwa kwa muda mrefu ili kupima implicit numerical diffusion ya mbinu. Katika tatizo hili laini lenye exact solution inayojulikana, kuenea kwa vortex kwa muda au kupungua kwa peak value kunaweza kutafsiriwa moja kwa moja kama kiashiria cha numerical diffusion.

Mean flow ilifafanuliwa kama:

\[ \rho=1,\qquad p=1,\qquad u=U_\infty,\qquad v=V_\infty \]

na kwa stationary case \(U_\infty=V_\infty=0\) ilichukuliwa.

Vortex velocity perturbations zilitolewa kama:

\[ (\delta u,\delta v)= \frac{\beta}{2\pi} e^{0{,}5(1-r^2)} (-\bar{y},\bar{x}) \]

na temperature perturbation kama:

\[ \delta T= -\frac{(\gamma-1)\beta^2} {8\gamma\pi^2} e^{1-r^2} \]

Entropy perturbation ni zero.

Kielelezo 4 kinaonyesha centerline density profiles katika nyakati tatu zikikaribia kuingiliana kabisa. Watafiti walitafsiri hili kama ushahidi kwamba mbinu haikuharibu vortex kwa implicit diffusion katika time units 100.

Ingawa matokeo haya ni muhimu, utafiti hautoi \(L_1\), \(L_2\) au \(L_\infty\) error norms. Ku-overlap kwa curves kwa kuona kunaonyesha physical structure imehifadhiwa, lakini hakutoi exact magnitude ya numerical error.

Jaribio la pili: Mach 7,5 hypersonic shear layer

Mfano wa pili ni two-dimensional mixing layer inayokua kati ya flows mbili zenye speeds na densities tofauti. Inlet velocity profile ina hyperbolic-tangent form:

\[ u(y)= \frac{U_1+U_2}{2} +\frac{U_1-U_2}{2} \tanh\left(\frac{2y}{\delta_\omega}\right). \]

Transition ilianzishwa kwa wall-normal velocity perturbations:

\[ v'= \sum_{k=1}^{2} a_k \cos\left(\frac{2\pi kt}{T}+\phi_k\right) \exp\left(-\frac{y^2}{b}\right). \]

Wavelength ilichaguliwa kuwa \(\lambda=30\), amplitudes za modes mbili \(a_1=a_2=0{,}05\), phases \(\phi_1=0\) na \(\phi_2=\pi/2\), na width parameter \(b=10\).

Convective Mach number ya 1,2 iliunda regime inayofaa kwa small shock structures kutokea ndani ya shear layer. Explicit artificial viscosity ilitumika katika test hii, na coefficient ilipunguzwa kadiri grid resolution ilivyoongezeka.

Kielelezo 5 kinaonyesha maendeleo ya Kelvin–Helmholtz instability kupitia density field:

1. Initial perturbations hukua katika streamwise direction.
2. Shear layer huwa wavy na kubadilika kuwa rolling vortices.
3. Vortices zilizo karibu huungana kwa jozi na kuunda structures kubwa zaidi.
4. Mixing-layer thickness huongezeka downstream.

Matokeo haya yanaonyesha kwamba mbinu inaweza kuwakilisha sharp shock-like changes na low-amplitude physical instabilities katika calculation moja. Hata hivyo, utafiti hautoi quantitative error comparison dhidi ya reference solution kwa vortex growth rate, mixing-layer thickness au dominant frequency.

Jaribio la tatu: Mach 6 boundary-layer transition

Jaribio la tatu linachunguza jinsi three-dimensional hypersonic boundary layer juu ya flat surface inavyobadilika kutoka instability waves hadi transition structures. Tatizo limetegemezwa kwenye previously published Mach 6 reference configuration.

Source inaeleza case ya \(T_w/T_\infty=6{,}5\) kama strongly cooled wall relative to hypersonic-flight conditions. Hapa neno “cooled” linapaswa kutafsiriwa relative to adiabatic au recovery temperature ambayo inaweza kuwa juu zaidi, si relative to free-stream static temperature.

Ili kuamsha transition, velocity forcing ya suction-blowing form ilitumika kwenye surface:

\[ v_w=f(x)g(z) \sum_{i=1}^{2} A_i\sin(\omega_it-\beta_iz). \]

Forcing region ni:

\[ 15\leq\frac{x-x_0}{\delta_0^*}\leq20 \]

Katika streamwise direction, center ya Gaussian envelope ilichaguliwa kuwa \(17{,}5\delta_0^*+x_0\), na width yake \(0{,}75\delta_0^*\). Kwa oblique modes mbili:

• Amplitude \(A=0{,}05U_\infty\),
• Frequency \(\omega=0{,}9c_\infty/\delta_0^*\),
• Spanwise wavenumbers \(\beta=\pm0{,}3/\delta_0^*\)

zilitolewa.

Kielelezo 6 kinaonyesha normalized density gradients katika plane parallel to wall na Q-criterion isosurfaces. Katika sehemu ya kwanza ya image kuna structures zilizopangwa na karibu na streamwise direction; katika sehemu inayofuata, vortex structures huongezeka, kupinda na kubadilika kuwa complex three-dimensional network.

Watafiti walihusisha maendeleo haya na mechanisms zifuatazo:

• Kukua kwa oblique instability waves,
• Nonlinear mode interactions,
• Uundaji wa streamwise vortices,
• Kutokea kwa three-dimensional transition structures,
• Kuanza kwa early stage ya breakdown.

Utafiti hauchunguzi quantitatively statistical properties za fully developed turbulence au surface-heat-flux overshoot. Matokeo yaliyoonyeshwa ni qualitative validation kwamba mbinu inaweza kutoa fundamental flow structures za transition.

Jaribio la nne: Mach 6 compression corner

Jaribio la mwisho ni three-dimensional flow field yenye compression ramp ya 8° baada ya flat plate. Mwanzo wa ramp umezungushwa kwa radius ndogo. Configuration hii ni classic high-speed aerodynamic problem ambapo shock-wave/boundary-layer interaction na strong thermal gradients karibu na wall zipo pamoja.

Inlet condition ilitengenezwa kutoka separate flat-plate precursor solution. Precursor field hii hutoa three-dimensional disturbances kwa inlet kwa kuendelea. Ili kupunguza excessive initial transient behavior, interior field ilianzishwa kwa smooth boundary-layer-like profile inayolingana na pressure.

Kielelezo 7 kinaonyesha temperature field katika symmetry plane. Thermal layer karibu na wall hukua juu ya flat plate. Flow inapofika kwenye corner na ramp, temperature karibu na wall huongezeka kutokana na compression. Downstream kwenye ramp, hot layer huwa nene na strong temperature gradients hujikusanya karibu na surface.

Test hii inaonyesha kwamba mbinu inaweza kuunda qualitative features zifuatazo katika complex geometry:

• Compression effect inayotokana na ramp,
• Temperature rise karibu na wall,
• Distortion ya thermal boundary layer,
• Kuendelea kwa strong gradients baada ya corner,
• Transport ya three-dimensional inlet disturbances downstream.

Hata hivyo, utafiti hautoi quantitative comparison dhidi ya experiment au high-resolution reference kwa surface pressure coefficient, Stanton number, heat flux, separation length au shock location.

Nguvu za utafiti

• Haar wavelet method imetumika kwa systematic namna kwenye three-dimensional compressible Navier–Stokes equations.
• Uwezo wa kuwakilisha discontinuities bila Gibbs oscillations umeonyeshwa moja kwa moja kwa Heaviside function.
• Validation problems zimeongezwa ugumu hatua kwa hatua kutoka smooth stationary vortex hadi three-dimensional hypersonic shock/boundary-layer interaction.
• Long-time vortex test inaonyesha kwamba low-amplitude structures hazikandamizwi na implicit diffusion.
• Kelvin–Helmholtz vortex formation na pairing zimetolewa katika Mach 7,5 shear layer.
• Oblique-mode growth na streamwise-vortex structures zimepatikana katika Mach 6 boundary layer.
• Strong near-wall thermal gradients zimetatuliwa katika Mach 6 compression corner.
• Hierarchical na sparse structure ya Haar matrices inatoa clear algorithmic path kwa future acceleration.
• Author contributions, conflict of interest, data access na funding information zimetolewa wazi.

Mapungufu ya utafiti

• Utafiti ni preprint ambayo haijapitia peer review.
• Hakuna validation kwa physical wind tunnel au flight experiment.
• Sehemu kubwa ya tests nne inategemea qualitative visual flow-structure assessment.
• Error norms, observed order of convergence na comprehensive grid-convergence tables hazijatolewa.
• Hakuna direct accuracy-cost comparison na WENO, TENO, compact difference au discontinuous Galerkin methods kwenye grid na time step zilezile.
• Explicit artificial viscosity ilitumika katika strong-gradient tests, lakini coefficients zote na sensitivity results hazijaripotiwa kwa kina.
• Current dense-matrix implementation ina cost ya \(O(N_i^2)\) na limited parallel efficiency.
• Proposed fast \(O(N_i)\) algorithm haikutekelezwa na performance yake haikupimwa.
• Adaptive mesh au real-time multiresolution activation haikutekelezwa katika utafiti huu.
• Hakuna validation kwenye complex industrial geometries au unstructured meshes.
• Chemical reaction, real-gas effects, vibrational energy na high-temperature gas physics hazikujumuishwa katika model.
• Data zitapatikana kwa request pekee; hakuna open-code repository wala reproducibility package iliyotolewa.

Utafiti unaunga mkono nini?

• Haar basis inaweza kuzuia Gibbs oscillations katika step-type discontinuities.
• Haar collocation method inaweza kuhifadhi smooth flow structures kwa muda mrefu kwa low implicit damping.
• Mbinu inaweza kukokotoa kwa uthabiti mifano yenye shocks au strong gradients pamoja na transition waves.
• Kelvin–Helmholtz vortices na hypersonic boundary-layer transition structures zinaweza kutengenezwa katika resolutions zilizochaguliwa.
• Regular sparsity structure ya Haar matrices inaweza kutumika kwa fast algorithms.
• Mbinu ni alternative inayostahili kuchunguzwa kwa hypersonic-transition calculations.

Utafiti hauthibitishi nini?

• Haithibitishi kwamba Haar method kwa ujumla ni more accurate kuliko WENO, TENO au methods nyingine za kisasa.
• Hauonyeshi kwamba mbinu ina lower computational cost kwa accuracy ileile.
• Haithibitishi kwamba itafanya kazi kwa stability katika shock strength, geometry au boundary condition yoyote.
• Hauonyeshi experimentally kwamba actual implementation itafikia \(O(N)\) speed wakati fast wavelet transform inapotumika.
• Hauonyeshi kwamba adaptive resolution inafanya kazi automatically na reliably.
• Hauthibitishi hypersonic regimes zenye real gas, chemical reaction au plasma.
• Haukokotoi aerodynamic, thermal au flight-safety performance ya specific aircraft.
• Hautoi solver iliyo tayari kwa industrial design au certification.

Ni uthibitishaji gani unahitajika baadaye?

Ili mbinu ipige hatua kutoka research level hadi general-purpose hypersonic solver level, kazi zifuatazo ni muhimu:

• Systematic \(L_1\), \(L_2\) na \(L_\infty\) error measurement kwenye known analytical solutions,
• Calculation ya observed convergence order katika different resolutions,
• Investigation ya athari ya artificial-viscosity coefficient kwenye shock thickness na transition waves,
• Comparison na WENO/TENO methods chini ya grid na CFL conditions zilezile,
• Measurement ya CPU/GPU times, memory consumption na parallel scaling,
• Implementation ya fast Haar transforms,
• Development ya adaptive multiresolution grid strategy,
• Validation dhidi ya wind-tunnel pressure, heat-flux na transition-location data,
• Extension kwa real gas na high-temperature thermochemistry,
• Application kwa complex three-dimensional vehicle geometries.

Mbinu na Matokeo ya Utafiti

Muundo wa kimetodolojia

Muhtasari wa kulinganisha majaribio manne

Maana ya kisayansi ya vielelezo

Matokeo makuu

• Haar basis haikutoa Gibbs-type numerical oscillations katika discontinuous test function ya utafiti.
• Stationary isentropic vortex ilihifadhiwa kwa time units 100 bila visible profile loss.
• Physical-instability growth, vortex roll-up na vortex pairing zilikokotolewa katika mixing layer yenye Mach 7,5 stream.
• Nonlinear interaction ya oblique modes na formation ya streamwise vortices zilizalishwa upya katika Mach 6 boundary layer.
• Near-wall temperature rise na strong thermal gradients zilitambuliwa katika Mach 6 compression ramp.
• Mbinu ilistabilishwa kwa explicit artificial viscosity katika strong gradients, na viscosity coefficient ikapunguzwa kadiri grid ilivyoboreshwa.
• Current dense-matrix implementation ililenga physical accuracy; high computational performance bado haijaonyeshwa.
• Structure ya Haar matrices ilionekana kufaa kwa future fast transforms na adaptive resolution.

Tafsiri ya makini ya matokeo

Utafiti unaonyesha kwamba Haar collocation approach inaweza kutumika kwa flows zenye hypersonic transition na strong gradients. Neno “inaweza kutumika” hapa linamaanisha kwamba tests zilizochaguliwa ziliendeshwa kwa uthabiti na fundamental physical structures zinazotarajiwa zikatolewa.

Matokeo bado hayaonyeshi kwamba mbinu kwa ujumla ni superior, faster au more accurate. Hilo linahitaji quantitative error, computational-time na memory comparison na current methods katika equal resolution. Pia, transition na compression-corner results zinahitaji kupimwa dhidi ya experimental pressure, heat-flux au transition-location measurements.

Maelezo ya Chanzo na Mbinu

Jina kamili la asili la utafiti: A Haar wavelet collocation method for hypersonic flows with strong gradients

Waandishi: Yu Zhang, Diego Exposito-Brioso na Xuerui Mao.

Majina yaliyofupishwa yaliyotumika katika utafiti: Y. Zhang, D. Exposito na X. Mao.

Mpangilio wa waandishi: Umehifadhiwa kwa mpangilio uliotolewa katika chanzo.

Mwandishi wa kwanza sawa: Hakuna taarifa ya equal contribution au equal first authorship.

Waandishi wa mawasiliano: Diego Exposito-Brioso na Xuerui Mao.

Anwani za mawasiliano: dieexbr17@gmail.com na xmao@bit.edu.cn.

Taasisi 1: School of Interdisciplinary Science, Beijing Institute of Technology, Beijing 100081, China.

Taasisi 2: School of Mechatronical Engineering, Beijing Institute of Technology, Beijing 100081, China.

Taasisi 3: Beijing Institute of Technology (Zhuhai), Zhuhai 519088, China.

Taasisi 4: State Key Laboratory of Explosion Science and Safety Protection, Beijing Institute of Technology, Beijing 100081, China.

DOI:10.2139/ssrn.7197528

Kiungo rasmi cha chanzo:Ukurasa rasmi wa rekodi ya SSRN

Jukwaa la uchapishaji: SSRN.

Mwaka wa uchapishaji: 2026.

Jarida: Hakuna jina maalumu la peer-reviewed journal au journal acceptance katika toleo hili.

Mchapishaji wa jarida lililopitiwa kitaalamu: Hakuna taarifa ya peer-reviewed journal publisher katika toleo hili.

Aina ya chanzo: Makala ya utafiti ya preprint yenye numerical-method development na validation kwa three-dimensional compressible flows.

Hali ya mapitio ya kitaalamu: Utafiti haujapitia peer review. Matokeo yanapaswa kutathminiwa kwa kuzingatia hali ya preprint na validation limitations zilizotajwa.

Michango ya Yu Zhang: Methodology, software, validation, formal analysis, investigation, data curation, original-draft writing na visualization.

Michango ya Diego Exposito-Brioso: Conceptualization, methodology, formal analysis, review na editing, supervision na funding acquisition.

Michango ya Xuerui Mao: Resources, review na editing, supervision, project administration na funding acquisition.

Ufadhili: National Natural Science Foundation of China, Grant No. W2433020.

Mgongano wa maslahi: Waandishi wametangaza kwamba hakuna known financial interest au personal relationship inayoweza kuathiri utafiti.

Upatikanaji wa data: Imeelezwa kwamba data zitatolewa kwa ombi. Hakuna open-data repository link iliyotolewa.

Upatikanaji wa msimbo: Hakuna open-source code, version tag au reproducibility package iliyowasilishwa.

Kikomo cha kimetodolojia: Utafiti unategemea calorically perfect gas assumption. Real gas, thermochemical nonequilibrium, chemical reaction, ionization na radiation hazijamodeliwa. Validation ya hypersonic cases inategemea kwa kiasi kikubwa qualitative reproduction ya expected flow structures.

Kikomo cha computational: Current implementation ina takribani \(O(N_i^2)\) cost kutokana na dense matrix operations. Fast \(O(N_i)\) wavelet algorithms na adaptive multiresolution implementation si matokeo yaliyotekelezwa, bali ni mapendekezo ya future work.

Onyo la artificial viscosity: Ingawa mbinu inadai kutotoa implicit numerical damping, imetumia explicit artificial viscosity katika strong shock interactions. Kwa hiyo, matokeo hayapaswi kutafsiriwa kama “numerical solution isiyo na viscosity kabisa”.

Maudhui haya ya Kituruki yameandaliwa kwa msingi wa maandishi, milinganyo, vielelezo, viambatisho na matokeo ya utafiti uliopakiwa. Hakuna dai la experimental success, flight performance, vehicle design, industrial superiority au proven general superiority dhidi ya mbinu nyingine ambalo halipo katika utafiti lililoongezwa.