Past exam of the mathematics course of the University of Cambridge / 2006 / iii / Paper 82 / 4 / b

For this sign of the viscous Burgers equation, the Cole-Hopf transformation is $q=2\delta {\partial }_{\theta }\log \psi$. Integrating the initial sinusoid gives $\psi (\theta ,0)=C\exp [-\cos \theta /(2\delta )]$; the positive multiplicative constant cancels from $q$. Set $a=1/(2\delta )$ and $s=\delta Z$. The Fourier series of the initial exponential and the heat equation yield the exact sinusoidal Cole-Hopf solution for negative-flux Burgers flow:Each harmonic decays by its heat multiplier, which can also be obtained directly by convolving it with the Gaussian heat-kernel convolution. At $s\gg 1$ the first harmonic dominates:The error bounds use $|I_n(a)|\le I_0(a)$ for real positive $a$, also immediate from the Fourier-integral representation. Dividing, including the correction from the denominator, givesFor $\delta \ll 1$, the large-argument asymptotic expansion of a modified Bessel function gives $I_1(a)/I_0(a)\to 1$. HenceThe sign is positive because the first Fourier harmonic of $\psi$ is negative, so its angular derivative is positive. For a fixed nonzero $\sin \theta$, the relative corrections vanish as $\delta \to 0$ and $s\to \infty$; at the symmetry points both the exact solution and the leading expression vanish.

Put $s=\delta Z>0$. For the Fay solution, $1/\sinh (ns)=2e^{-ns}[1+O(e^{-2ns})]$. At $s\gg 1$ the first harmonic therefore giveswhich agrees with the late-time limit in (i).

For the opposite limit, expand the reciprocal hyperbolic sine as a geometric series and interchange absolutely convergent sums at every $s>0$:This exact positive-denominator form is useful near the shock. For small $s$ and $|\theta |$, expand the denominator of the terms with small $(2j+1)s$ and use $\sin \theta \sim \theta$. The leading sum isFor the shock-layer scaling $\theta =s\eta$, the passage to this sum follows from dominated convergence: $\cosh z-1\ge z^2/2$ and $1-\cos \theta \ge c{\theta }^2$ for small $|\theta |$ provide a summable bound proportional to $[(2j+1{)}^2+{\eta }^2{]}^{-1}$. The same leading expression matches the outer small-angle range $s\ll |\theta |\ll 1$. More generally, split the exact sum at $(2j+1)s=b$ with $\max (s,|\theta |)\ll b\ll 1$: the low terms have vanishing relative Taylor error, while the discarded exact and approximate tails are smaller than the leading expression by $O(\max (s,|\theta |)/b)$.

Now apply the partial-fraction identity for the hyperbolic tangent with $\eta =\theta /s$:The Fay shock-layer asymptotics are thereforeThe angular shock thickness is $O(\delta Z)$, and its matched states are $\pm \pi /Z$. At $\theta =0$, both formulas vanish; the limiting slopes agree. The denominator throughout this derivation is $\sinh (n\delta Z)$ as printed in the PDF, not the erroneous $\sin (n\delta Z)$ in the converted TeX.