How the green function for the relativistic heat equation converges to the green function of the heat equation?

67 Views Asked by At

The relativistic heat equation or telegraphers equation is: $$ (\alpha\partial_t^2 + \beta\partial_t - \omega\,\nabla^2_{\text{3D}})G_R = \delta $$

if $\alpha \rightarrow 0$ the solution must converge to the non-relativistic heat equation $$ (\beta\partial_t - \omega\,\nabla^2_{\text{3D}})G = \delta $$

From Wikipedia https://en.m.wikipedia.org/wiki/Green%27s_function#Table_of_Green%27s_functions the green function for the 3D relativistic heat equation with $\gamma = \frac{\beta}{2\alpha}$ and $c = \sqrt{\frac{\omega}{\alpha}}$ is: $$ (\partial_t^2 + \frac{\beta}{\alpha}\partial_t - \frac{\omega}{\alpha}\,\nabla^2_{\text{3D}})G_R = \frac{1}{\alpha}\delta $$ $$ (\partial_t^2 + 2\gamma\partial_t - c^2\,\nabla^2_{\text{3D}}) G'_R = \delta $$ so $G_R = \frac{G'_R}{\alpha}$ $$ G'_R =\frac{e^{-\gamma t}}{20\pi} \left[\left(8-3e^{-\gamma t}+2\gamma t+4\gamma^2t^2\right)\frac{\delta(ct-r)}{r^2}+\frac{\gamma^2}{c}\Theta(ct - r)\left(\frac{1}{cu}I_1\left(\frac{\gamma u}{c}\right)+\frac{4 t}{u^2}I_2\left(\frac{\gamma u}{c}\right)\right)\right] $$

with $ u=\sqrt{c^2t^2-r^2}$

if $\alpha \rightarrow 0$ then $\gamma, c \rightarrow \infty$ so: $$ G'_R =\frac{e^{-\gamma t}}{20\pi} \left[\frac{\gamma^2}{c}\Theta(t)\left(\frac{1}{cu}I_1\left(\frac{\gamma c t}{c}\right)+\frac{4 t}{c^2 t ^2}I_2\left(\frac{\gamma c t}{c}\right)\right)\right] = \frac{e^{-\gamma t}}{20\pi} \left[\frac{\gamma^2}{c}\Theta(t)\left(\frac{1}{c^2 t}I_1\left(\frac{\gamma c t}{c}\right)+\frac{4 t}{c^2 t ^2}I_2\left(\frac{\gamma u}{c}\right)\right)\right] $$ The modified bessle function according to https://personal.math.ubc.ca/~cbm/aands/abramowitz_and_stegun.pdf is asymptotic to $I_\nu(z) = \frac{e^z}{\sqrt{2\pi z}}$ $$ G'_R =\frac{\exp{(-\gamma t)}}{20\pi} 5 \left[\frac{\gamma^2}{c^3 t}\Theta(t)\left(\frac{exp(\gamma t))}{\sqrt{2\pi \gamma t}})\right)\right] exp(-\frac{\gamma r^2}{2 c^2 t})=\frac{1}{4\pi} \left[\frac{\gamma^2}{c^3 t}\frac{1}{\sqrt{2\pi \gamma t}}\right]\Theta(t)exp(-\frac{\gamma r^2}{2 c^2 t}) $$

Now substituting in the equation $\alpha$, $\beta$, $\omega$ we get: $$ G_R = \frac{G'_R}{\alpha} = \frac{1}{4\pi \alpha} \left[\frac{(\frac{\beta}{2\alpha})^2}{(\frac{\omega}{\alpha})^\frac{3}{2} t}\frac{1}{\sqrt{2\pi \frac{\beta}{2\alpha} t}}\right]\Theta(t) exp(-\frac{\beta r^2}{4 \omega t}) = \frac{1}{4\pi} \left[\frac{(\frac{\beta}{2})^2}{(\omega)^\frac{3}{2} t}\frac{1}{\sqrt{2\pi \frac{\beta}{2} t}}\right]\Theta(t) \frac{\alpha^\frac{3}{2}\alpha^\frac{1}{2}}{\alpha^3}exp(-\frac{\beta r^2}{4 \omega t}) $$ Therefore:

$$ G_R = \frac{1}{4\pi} \left[\frac{(\frac{\beta}{2})^2}{(\omega)^\frac{3}{2} t}\frac{1}{\sqrt{2\pi \frac{\beta}{2} t}}\right]exp(-\frac{\beta r^2}{4 \omega t})\Theta(t) \frac{1}{\alpha} = \frac{\beta^\frac{3}{2}}{16(\omega \pi t)^\frac{3}{2}}exp(-\frac{\beta r^2}{4 \omega t})\Theta(t) \frac{1}{\alpha} $$ the solution is approximately correct to the non-relativistic one in the dependence of time and space but there are issues with the constants and it must not depend on $\alpha$ the non-relativistic approximation.

The non-relativistic green function of the heat equation is: $$ G = \frac{\beta^\frac{1}{2}}{(4 \pi \omega t)^{\frac{3}{2}}}exp(-\frac{r^2 \beta}{4 \omega t}) $$

I do not understand if there is something wrong with the solution found on Wikipedia or I am doing some errors in the approximations