Normal shock wave

A normal shock is a discontinuity perpendicular to the flow direction across which mass, momentum, and total energy fluxes are conserved.

Property ratios for this discontinuity are tabulated in normal shock tables.

A tube in which releasing an initial pressure difference produces a shock, often accompanied by an expansion wave and a contact surface. Spatially separated pressure probes can measure shock arrival speed and jump size. The shock and the contact surface must be identified separately when interpreting the downstream state.

For a perfect-gas normal shock wave reflected from a stationary rigid wall, apply the Rankine-Hugoniot conditions for a perfect gas using signed velocity in each shock frame. The once-shocked state is the common left state for both applications. Its two relative speeds are roots of $v^2-(\gamma +1)u_{s}v/2-\gamma p_s/{\rho }_s=0$. Their product and the pressure jump eliminate the velocity, giving the displayed relation. A strong incident shock wave gives $p_1/p_s\to (3\gamma -1)/(\gamma -1)$.

For a moving level surface, $n=\nabla S/|\nabla S|$, $v_n=-S_t/|\nabla S|$, and the surface delta distribution is ${\delta }_s=|\nabla S|\delta (S)$. Piecewise classical conservation laws acquire a distributional flux defect $[b]\cdot n-v_n[a]$. Vanishing of this defect is the Rankine-Hugoniot condition.

Normal shock tables tabulate upstream-to-downstream property ratios for a normal shock wave, typically as functions of upstream Mach number and heat capacity ratio. They encode the Rankine-Hugoniot conditions for a perfect gas rather than defining the discontinuity itself.

The shock frame is an inertial reference frame in which the shock is stationary. A travelling shock then becomes a steady flow through a fixed discontinuity.

In the rest frame of a one-dimensional shock, states with velocities $w_0,w_1$ satisfyand

These are the Rankine-Hugoniot conditions specialized to the mass, momentum and energy fluxes of a perfect gas.

For a stationary normal shock in a polytropic perfect gas, the upstream and downstream normal speeds obey $u_1{u}_2=c_{\mathrm{cr}}^2$, where $c_{\mathrm{cr}}$ is the sonic speed on their common Bernoulli streamline.

When the upstream pressure is negligible, a perfect-gas shock has density ratio $(\gamma +1)/(\gamma -1)$. For $\gamma =5/3$, the downstream density is four times the upstream density.

A strong, nonradiative shock in a monatomic ideal gas gives $p_{\mathrm{d}}=3{\rho }_{\mathrm{u}}{v}_{\mathrm{u}}^2/4$ and ${\rho }_{\mathrm{d}}=4{\rho }_{\mathrm{u}}$. The ideal gas law therefore gives the temperature in the title, with $\mu$ the mean particle mass. If the upstream speed is a Newtonian free-fall speed, $k_{\mathrm{B}}{T}_{\mathrm{d}}=3\mu GM/(8R)$. Mass alone does not determine this temperature without a radius and composition; electron-ion disequilibrium and radiation can also make the emergent spectrum differ from the ideal gas temperature.

In the rest frame of a strong planar shock wave, a perfect gas has ${\rho }_2/{\rho }_1=(\gamma +1)/(\gamma -1)$, $u_2/u_1=(\gamma -1)/(\gamma +1)$ and $p_2=2{\rho }_1{u}_1^2/(\gamma +1)$. Its downstream specific internal energy is ${ϵ}_2=p_2/[(\gamma -1){\rho }_2]=2u_1^2/(\gamma +1{)}^2$. Multiplying by mass flux $j={\rho }_1{u}_1$ gives the internal-energy flux in the title. The lost bulk kinetic energy flux instead equals $j(u_1^2-u_2^2)/2=\gamma F_{\mathrm{int}}$: this includes downstream pressure work and is the increase of specific enthalpy flux in the cold-upstream limit. Neither quantity is a temperature change per unit time of a fixed target mass without further information about that target.

The shock compression ratio is the downstream-to-upstream density ratio. For a strong perfect-gas shock,In the upstream rest frame, the downstream velocity is $2V_s/(\gamma +1)$ and its pressure is $2{\rho }_0{V}_s^2/(\gamma +1)$.

A plane shock moving at speed $V_s$ through initially stationary gas has upstream speed $u_1=V_s$ and downstream speed $u_2=V_s/\chi$ in the shock frame, where $\chi$ is the shock compression ratio. Transforming back to the initial slab frame gives $v_2=V_s-u_2$. At the breakout time $L/V_s$, a uniform downstream slab has thickness $L-v_{2}L/V_s=L/\chi$. The strong monatomic limit has $\chi =4$, hence $L_2=L/4$ and $v_2=3V_s/4$. This uses a maintained planar post-shock state; boundary rarefactions and variations of shock speed can change the actual thickness of a freely evolving finite cloud.

For a strong monatomic shock, the planar shock compression of a finite slab gives bulk speed $3V_s/4$ and kinetic energy $E_{\mathrm{bulk}}=9M_c{V}_s^2/32$. The initial internal energy is $E_{\mathrm{th}}=3M_{c}p/(2\rho )$. A bulk-ballistic escape estimate $E_{\mathrm{bulk}}\gtrsim W$ uses positive binding-energy magnitude $W$, not negative gravitational potential energy. With adiabatic sound speed $c_{\mathrm{ad}}^2=5p/(3\rho )$ it yields the first displayed coefficient. Using the isothermal sound speed $c_T^2=p/\rho$ instead yields $16/3\simeq 5.3$. The definitions obey ${\mathcal{M}}_T^2=(5/3){\mathcal{M}}_{\mathrm{ad}}^2$. Heating, subsequent expansion and energy loss can alter the physical escape threshold, so the coefficient is not independent of the energy budget and Mach convention.

If a piston moves at speed $V$ into stationary perfect gas and drives a shock at speed $U$, then the shock-frame upstream and downstream speeds are $U$ and $U-V$. If $p_2-p_1=\beta p_1$, the jump conditions give

If a shock travels into stationary gas at speed $V$ while the downstream piston and gas travel at $V/3$, then for $\gamma =3$,

Let $P=p_2/p_1$ and $D={\rho }_2/{\rho }_1$ be the pressure and mass density ratios across a normal shock wave in a perfect gas with specific-heat ratio $\gamma$. Eliminating the velocities from the Rankine-Hugoniot conditions for a perfect gas givesAs $P\to \infty$, the ratio tends to the shock compression ratio $(\gamma +1)/(\gamma -1)$.

For the pressure-density Hugoniot relation for a perfect gas, the change in specific entropy is $[s]/c_v=\log P-\gamma \log D(P)$, with $c_v$ the specific heat capacity at constant volume. Its derivative isIt is positive for a compressive normal shock wave with $P>1$. For a weak shock with $P=1+\delta$, integrating the leading term givesThus the entropy production is cubic in the small pressure jump, even though the mass density and temperature changes already appear at first order.

For $N$ equal weak shocks with individual pressure ratio $1+\delta$ and total ratio $P_s=(1+\delta {)}^N$, entropy production in a perfect-gas shock givesAt fixed $P_s>1$, this tends to zero as the compression is divided into increasingly many smaller weak shocks. A single finite normal shock wave instead gives the positive value $\log P_s-\gamma \log D(P_s)$. Gradual compression can consequently approach reversible isentropic flow while an abrupt compression generates finite entropy.

A weak shock has a small relative pressure jump. Its density jump agrees with the reversible adiabatic prediction through second order in the pressure jump, while entropy production first appears at third order.

An oblique shock is inclined relative to the upstream velocity. In an inviscid perfect gas, the normal velocity satisfies the normal-shock jump conditions while the tangential velocity is continuous.

Measure shock angle $\beta$ from the incident flow to the shock front, and deflection $\theta$ towards it. The strong normal-compression ratio and continuity of tangential velocity give $\tan \beta =[(\gamma +1)/(\gamma -1)]\tan (\beta -\theta )$. Angle-addition identities reduce this to $\sin (2\beta -\theta )=\gamma \sin \theta$. The two title branches exist for $0<\theta <\arcsin (1/\gamma )$ and merge at $\beta =\pi /4+\theta /2$. As $\theta \to 0$, their formal limits are $(\gamma +1)\theta /2$ and $\pi /2-(\gamma -1)\theta /2$. The weak-branch expansion requires $M\sin \beta \gg 1$; at finite Mach number its true zero-deflection limit is the Mach angle.

A shock polar is the locus of possible downstream velocity vectors, or an equivalent pressure/deflection relation, as the orientation of an oblique shock varies for fixed upstream conditions. Entropy admissibility selects physical compressive shocks. The strong-shock perfect-gas velocity polar is a circle, while finite-Mach polars have a different shape.

The normal velocity is reduced by $q$ in the strong-shock limit while tangential velocity is unchanged. If the upstream speed is $U$ and the shock normal makes angle $\beta$ with its direction, then $u_X=U[1-(1-q){\cos }^2\beta ]$ and $u_Y=\pm U(1-q)\cos \beta \sin \beta$. Eliminating the angle gives the displayed circle, centered at $\gamma U/(\gamma +1)$ with radius $U/(\gamma +1)$. It is the limiting shock polar in downstream velocity space.

For the circle with center distance $a=\gamma U/(\gamma +1)$ and radius $R=U/(\gamma +1)$, the most deflected ray from the origin is tangent. The right triangle formed by the origin, center and tangency point gives $\sin {\delta }_{\max }=R/a=1/\gamma$. This is a velocity-vector deflection angle, not the angle of the shock normal.

If $p_2/p_1=1+ϵ$ with $0<ϵ\ll 1$ and the upstream Mach number is $M_1>1$, the shock angle is the Mach angle to leading order and the flow deflection isThe normal speed decreases, so the downstream flow turns toward the shock front.

The Mach angle of a supersonic flow with Mach number $M>1$ is $\mu =\arcsin (M^{-1})$. Infinitesimal pressure disturbances lie on the corresponding Mach cone, and a weak oblique shock approaches this angle as its strength tends to zero.

The Mach angle is the half-angle between a Mach cone generator and its backward motion axis for a steady point source in a uniform three-dimensional medium with isotropy and constant sound speed. It obeysThe relation concerns weak acoustic waves in linear acoustics. It is not the angle of a finite oblique shock produced by finite turning of the fluid flow; that angle also depends on the deflection and the gas properties.^[ref]

For a short proof, choose an old emission in the moving-point acoustic wavefronts construction and set $s=-\tau >0$. Its circle has center $(-Us,0)$ and radius $cs$. A tangent line from the present source at the origin meets the circle at a right angle to its radius. The resulting right triangle has hypotenuse $Us$ and opposite side $cs$, giving the displayed sine ratio. The emission age cancels: every wavefront has the same tangent directions, so a continuum of expanding spheres produces one simple Mach cone.

In an original example, stipulate uniform gas with sound speed $c=340\mathrm{m}{\mathrm{s}}^{-1}$ and a source moving through it at $U=680\mathrm{m}{\mathrm{s}}^{-1}$. Then $M=2$ and $\mu =\arcsin (1/2)=30^{\circ }$. An emission one second earlier is now a sphere centered $680\mathrm{m}$ behind the source with radius $340\mathrm{m}$. The selected sound speed defines this example; it is not a constant for air at every temperature. The Mach angle decreases as $M$ increases: its derivative is $-1/(M\sqrt{M^2-1})$ in radians, and its limits are $90^{\circ }$ as $M\downarrow 1$ and $0$ as $M\to \infty$.

The Mach cone forms the geometric envelope of acoustic wave wavefronts from a source moving at a supersonic Mach number $M>1$. In the source reference frame the ray-velocity sphere is centered at $-M{c}_0{\mathbf{e}}_x$ and has radius $c_0$. Its tangent rays make angle ${\sin }^{-1}(1/M)$ to the backward flight direction; all other emitted acoustic rays lie inside that Mach cone.

The Mach cone is the geometric envelope of the spheres in moving-point acoustic wavefronts. To obtain its axial boundary without a diagram, put $s=-\tau >0$ and impose the geometric envelope conditions $F=0$ and ${\partial }_{s}F=0$ onThe partial derivative condition gives $x=-(U^2-c^2)s/U$. Substituting into $F=0$ gives $y^2=c^2(U^2-c^2)s^2/U^2$, henceFor $M=2$, the Mach cone section has slope magnitude $1/\sqrt{3}$, consistent with a Mach angle of $30^{\circ }$. If emission begins at a finite time, the reached region is limited by the oldest wavefront; the infinite cone describes the steady geometric boundary obtained with an indefinitely long source history. In a uniformly moving gas, advection shifts each emission's sphere by the gas velocity times its age; working with the source's relative velocity through the gas restores the same construction.