betaFeedback

WorksPhysics and Astronomy

Novel distributional Laplacians and a Coulomb-gauge problem

Abstract

Novel distributional Laplacians of an arsinh function and a related logarithm function are derived. The method used to establish the former result is employed in a calculation of the gauge transformation function for the case of a uniformly moving point charge.

Introduction

The classical Laplacian of the function arsinh⁡[a(z−b)/s]\operatorname{arsinh}[a(z-b)/s], where s=(x2+y2)1/2s=(x^2+y^2)^{1/2} and zz are cylindrical coordinates, and aa and bb are arbitrary independent constants, is given by1

Δarsinh⁡a(z−b)s=a(1−a2)(z−b)[s2+a2(z−b)2]3/2.(1)\Delta\operatorname{arsinh}\frac{a(z-b)}{s}=\frac{a(1-a^2)(z-b)}{[s^2+a^2(z-b)^2]^{3/2}}. \tag*{(1)}

It may come as a surprise that the distributional (generalized) Laplacian adds to the classical one a delta-function term,

Δˉarsinh⁡a(z−b)s=a(1−a2)(z−b)[s2+a2(z−b)2]3/2−sgn⁡[a(z−b)]δ(s)s,(2)\bar{\Delta}\operatorname{arsinh}\frac{a(z-b)}{s}=\frac{a(1-a^2)(z-b)}{[s^2+a^2(z-b)^2]^{3/2}}-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}, \tag*{(2)}

where the bar in Δˉ\bar{\Delta} denotes it as a distributional operator, sgn⁡(⋅)\operatorname{sgn}(\cdot) is the sign function and the delta function δ(s)\delta(s) is normalized as ∫0∞ds δ(s)=1\int_{0}^{\infty}\mathrm{d}s\,\delta(s)=1. Several novel distributional Laplacians involving the arsinh function can be obtained with specific values of the constants in Eq. (2). For example, when a=1a=1 and b=0b=0, we obtain

Δˉarsinh⁡zs=−sgn⁡(z)δ(s)s,(3)\bar{\Delta}\operatorname{arsinh}\frac{z}{s}=-\operatorname{sgn}(z)\frac{\delta(s)}{s}, \tag*{(3)}

which is similar to the distributional Laplacian of the logarithm function,

Δˉln⁡(s)=δ(s)s,(4)\bar{\Delta}\ln(s)=\frac{\delta(s)}{s}, \tag*{(4)}

an informal proof of which was given in [1]. The generation of an additional delta-function term by the application of the distributional Laplacian in cylindrical coordinates to the function arsinh⁡[a(z−b)/s]\operatorname{arsinh}[a(z-b)/s] is similar to the distributional Laplacian in spherical coordinates generating additional distributional terms when it is applied to some functions [2, 3].

In § 2, we give an informal proof of the distributional Laplacian (2); another, related, distributional Laplacian is there derived also. In § 3, the method used to establish the mathematical result (2) is employed in a calculation of the gauge function of the transformation of the Lorenz-gauge potentials of a uniformly moving charge to the Coulomb gauge. The last section contains some concluding remarks.

Informal proofs

The distributional relation (2) can be established by solving the (distributional) Poisson equation

Δˉg(s,z)=a(1−a2)(z−b)[s2+a2(z−b)2]3/2−sgn⁡[a(z−b)]δ(s)s,(5)\bar{\Delta}g(s,z)=\frac{a(1-a^{2})(z-b)}{\left[s^{2}+a^{2}(z-b)^{2}\right]^{3/2}}-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}, \tag*{(5)}

using the standard integral representation of its solution, the operator of which we term the inverse Laplacian and denote by Δˉ−1\bar{\Delta}^{-1}. Thus

g(s,z)=Δˉ−1[a(1−a2)(z−b)[s2+a2(z−b)2]3/2−sgn⁡[a(z−b)]δ(s)s]=−14π∫d3r′∣r−r′∣[a(1−a2)(z′−b)[s′2+a2(z′−b)2]3/2−sgn⁡[a(z′−b)]δ(s′)s′].(6)\begin{aligned} g(s,z)&=\bar{\Delta}^{-1}\left[\frac{a(1-a^{2})(z-b)}{\left[s^{2}+a^{2}(z-b)^{2}\right]^{3/2}}-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}\right]\\ &=-\frac{1}{4\pi}\int\frac{\mathrm{d}^{3}r'}{|\boldsymbol{r}-\boldsymbol{r}'|}\left[\frac{a(1-a^{2})(z'-b)}{\left[s'^{2}+a^{2}(z'-b)^{2}\right]^{3/2}}-\operatorname{sgn}[a(z'-b)]\frac{\delta(s')}{s'}\right]. \tag*{(6)} \end{aligned}

Since we use cylindrical coordinates, we employ in (6) a cylindrical-coordinate expansion of the inverse distance ([4], p 140),

1∣r−r′∣=∑m=−∞∞∫0∞dk eim(ϕ−ϕ′)Jm(ks)Jm(ks′)e−k(z>−z<),(7)\frac{1}{|\boldsymbol{r}-\boldsymbol{r}'|}=\sum_{m=-\infty}^{\infty}\int_{0}^{\infty}\mathrm{d}k\,e^{im(\phi-\phi')}J_m(ks)J_m(ks')e^{-k(z_{>}-z_{<})}, \tag*{(7)}

where Jm(⋅)J_m(\cdot) are the Bessel functions of the first kind of order mm and z>z_{>} (z<z_{<}) is the greater (lesser) of the zz-coordinates of the vectors r\boldsymbol{r} and r′\boldsymbol{r}'.

We calculate first the inverse Laplacian of the non-delta-function part of the RHS of (5),

Δˉ−1a(1−a2)(z−b)[s2+a2(z−b)2]3/2=a(a2−1)2∫0∞dk J0(ks)∫−∞∞dz′(z′−b)e−k(z>−z<)∫0∞s′ ds′ J0(ks′)[s′2+a2(z′−b)2]3/2=a(a2−1)2∣a∣∫0∞dk J0(ks)∫−∞∞dz′sgn⁡(z′−b)e−k(z>−z<)e−∣a(z′−b)∣k,(8)\begin{aligned} \bar{\Delta}^{-1}\frac{a(1-a^{2})(z-b)}{\left[s^{2}+a^{2}(z-b)^{2}\right]^{3/2}} &=\frac{a(a^{2}-1)}{2}\int_{0}^{\infty}\mathrm{d}k\,J_{0}(ks)\int_{-\infty}^{\infty}\mathrm{d}z'(z'-b)e^{-k(z_{>}-z_{<})}\int_{0}^{\infty}s'\,\mathrm{d}s'\,\frac{J_{0}(ks')}{\left[s'^{2}+a^{2}(z'-b)^{2}\right]^{3/2}}\\ &=\frac{a(a^{2}-1)}{2|a|}\int_{0}^{\infty}\mathrm{d}k\,J_{0}(ks)\int_{-\infty}^{\infty}\mathrm{d}z'\operatorname{sgn}(z'-b)e^{-k(z_{>}-z_{<})}e^{-|a(z'-b)|k}, \tag*{(8)} \end{aligned}

where the 2nd line is simplified as a result of ∫02πdϕ′exp⁡[im(ϕ−ϕ′)]=2πδm0\int_{0}^{2\pi} d\phi' \exp[im(\phi-\phi')] = 2\pi\delta_{m0} and the 3rd line employs the result

∫0∞s′ ds′ J0(ks′)[s′2+a2(z′−b)2]3/2=1∣a(z′−b)∣e−∣a(z′−b)∣k,(9)\int_{0}^{\infty} s'\,ds'\,\frac{J_0(ks')}{[s'^2+a^2(z'-b)^2]^{3/2}}=\frac{1}{\lvert a(z'-b)\rvert}e^{-\lvert a(z'-b)\rvert k}, \tag*{(9)}

obtained using a tabulated integral ([5], item 2.12.4(28)). The evaluation of the integral with respect to z′z' in (8) is cumbersome, but straightforward. It yields

∫−∞∞dz′ sgn⁡(z′−b)e−k(z>−z<)e−∣a(z′−b)∣k=∫−∞zdz′ sgn⁡(z′−b)e−k(z−z′)e−∣a(z′−b)∣k+∫z∞dz′ sgn⁡(z′−b)e−k(z′−z)e−∣a(z′−b)∣k=2(a2−1)k{1−e−∣a∣(z−b)k−(1−e−(z−b)k),z>b,1−e−(b−z)k−(1−e−∣a∣(b−z)k),z<b.(10)\begin{aligned} \int_{-\infty}^{\infty} dz'\,\operatorname{sgn}(z'-b)e^{-k(z_>-z_<)}e^{-\lvert a(z'-b)\rvert k} ={}&\int_{-\infty}^{z} dz'\,\operatorname{sgn}(z'-b)e^{-k(z-z')}e^{-\lvert a(z'-b)\rvert k}\\ &+\int_{z}^{\infty} dz'\,\operatorname{sgn}(z'-b)e^{-k(z'-z)}e^{-\lvert a(z'-b)\rvert k}\\ ={}&\frac{2}{(a^2-1)k} \begin{cases} 1-e^{-\lvert a\rvert(z-b)k}-(1-e^{-(z-b)k}), & z>b,\\ 1-e^{-(b-z)k}-(1-e^{-\lvert a\rvert(b-z)k}), & z<b. \end{cases} \tag*{(10)} \end{aligned}

The peculiar writing of this result is to facilitate the use of a tabulated integral ([5], item 2.12.8(5)),

∫0∞dk 1k(1−e−ρk)J0(ks)=arsinh⁡ρs,(11)\int_{0}^{\infty} dk\,\frac{1}{k}(1-e^{-\rho k})J_0(ks)=\operatorname{arsinh}\frac{\rho}{s}, \tag*{(11)}

in the remaining integration in (8), which gives

Δˉ−1a(1−a2)(z−b)[s2+a2(z−b)2]3/2=sgn⁡(a)[arsinh⁡∣a∣(z−b)s−arsinh⁡z−bs]=arsinh⁡a(z−b)s−sgn⁡(a)arsinh⁡z−bs.(12)\begin{aligned} \bar{\Delta}^{-1}\frac{a(1-a^2)(z-b)}{[s^2+a^2(z-b)^2]^{3/2}} &=\operatorname{sgn}(a)\left[\operatorname{arsinh}\frac{\lvert a\rvert(z-b)}{s}-\operatorname{arsinh}\frac{z-b}{s}\right]\\ &=\operatorname{arsinh}\frac{a(z-b)}{s}-\operatorname{sgn}(a)\operatorname{arsinh}\frac{z-b}{s}. \tag*{(12)} \end{aligned}

Since arsinh⁡(−x)=−arsinh⁡(x)\operatorname{arsinh}(-x)=-\operatorname{arsinh}(x), this result holds true for both z>bz>b and z<bz<b.

We now calculate the inverse Laplacian of the delta-function part of (5),

Δˉ−1[−sgn⁡[a(z−b)]δ(s)s]=sgn⁡(a)2∫0∞dk J0(ks)∫−∞∞dz′ sgn⁡(z′−b)e−k(z>−z<)=sgn⁡(a)2∫0∞dk J0(ks){e−kz[−∫−∞bdz′ekz′+∫bzdz′ekz′]+ekz∫z∞dz′e−kz′}=sgn⁡(a)∫0∞dk J0(ks)1k[1−e−k(z−b)],(13)\begin{aligned} \bar{\Delta}^{-1}\left[-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}\right] ={}&\frac{\operatorname{sgn}(a)}{2}\int_{0}^{\infty}dk\,J_0(ks)\int_{-\infty}^{\infty}dz'\,\operatorname{sgn}(z'-b)e^{-k(z_>-z_<)}\\ ={}&\frac{\operatorname{sgn}(a)}{2}\int_{0}^{\infty}dk\,J_0(ks)\left\{e^{-kz}\left[-\int_{-\infty}^{b}dz'e^{kz'}+\int_{b}^{z}dz'e^{kz'}\right]+e^{kz}\int_{z}^{\infty}dz'e^{-kz'}\right\}\\ ={}&\operatorname{sgn}(a)\int_{0}^{\infty}dk\,J_0(ks)\frac{1}{k}[1-e^{-k(z-b)}], \tag*{(13)} \end{aligned}

where the 2nd line is simplified as a result of the integral with respect to ϕ′\phi' equaling 2πδm02\pi\delta_{m0} and of the integral with respect to s′s' yielding a unity factor J0(0)J_0(0) on account of the factor δ(s′)/s′\delta(s')/s'; in the 3rd line, z>bz>b is assumed. The last line of Eq. (13) can be evaluated using the integral (11), yielding

Δˉ−1[−sgn⁡[a(z−b)]δ(s)s]=sgn⁡(a)arsinh⁡z−bs,(14)\bar{\Delta}^{-1}\left[-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}\right]=\operatorname{sgn}(a)\operatorname{arsinh}\frac{z-b}{s}, \tag*{(14)}

which, as the result (12), holds true for both z>bz>b and z<bz<b. Adding now Eqs. (12) and (14) gives

Δˉ−1[a(1−a2)(z−b)[s2+a2(z−b)2]3/2−sgn⁡[a(z−b)]δ(s)s]=arsinh⁡a(z−b)s.(15)\bar{\Delta}^{-1}\left[\frac{a(1-a^2)(z-b)}{[s^2+a^2(z-b)^2]^{3/2}}-\operatorname{sgn}[a(z-b)]\frac{\delta(s)}{s}\right]=\operatorname{arsinh}\frac{a(z-b)}{s}. \tag*{(15)}

The Poisson equation (5) is thus solved by

g(s,z)=arsinh⁡a(z−b)s,(16)g(s,z)=\operatorname{arsinh}\frac{a(z-b)}{s}, \tag*{(16)}

which establishes the distributional-Laplacian relation (2) as holding true.

Using the identity arsinh⁡(x)=ln⁡(x+1+x2)\operatorname{arsinh}(x)=\ln(x+\sqrt{1+x^2}) and relation (4) for Δˉln⁡(s)\bar{\Delta}\ln(s), the distributional relation (2) can be written as

Δˉln⁡[a(z−b)+s2+a2(z−b)2]=a(1−a2)(z−b)[s2+a2(z−b)2]3/2+[1−sgn⁡[a(z−b)]]δ(s)s.(17)\bar{\Delta}\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2}\right]=\frac{a(1-a^2)(z-b)}{[s^2+a^2(z-b)^2]^{3/2}}+\left[1-\operatorname{sgn}[a(z-b)]\right]\frac{\delta(s)}{s}. \tag*{(17)}

Intriguingly, the distributional Laplacian of this logarithmic function generates a delta-function term only when a(z−b)a(z-b) is negative. Unfortunately, this result cannot be established independently of the validity of (2) and (4) by a calculation of the inverse Laplacian of the RHS of (17) since such calculation diverges due to the divergence of Δˉ−1[δ(s)/s]\bar{\Delta}^{-1}[\delta(s)/s]. But the distributional Laplacian (17) can be established as

Δˉln⁡[a(z−b)+s2+a2(z−b)2]=wlim⁡ϵ→0Δˉln⁡[a(z−b)+s2+a2(z−b)2+ϵ2],(18)\bar{\Delta}\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2}\right]=\underset{\epsilon\to0}{\operatorname{wlim}}\bar{\Delta}\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2+\epsilon^2}\right], \tag*{(18)}

where the symbol wlim denotes the weak limit (see, e.g., [6]), since it can be shown that

lim⁡ϵ→0∫0∞s dsf(s) Δln⁡[a(z−b)+s2+a2(z−b)2+ϵ2]=a(1−a2)(z−b)∫0∞s dsf(s)[s2+a2(z−b)2]3/2+[1−sgn⁡[a(z−b)]]f(0),(19)\begin{aligned} \lim_{\epsilon\to0}\int_0^\infty s\,\mathrm{d}s f(s)\,\Delta\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2+\epsilon^2}\right] &=a(1-a^2)(z-b)\int_0^\infty\frac{s\,\mathrm{d}s f(s)}{[s^2+a^2(z-b)^2]^{3/2}}\\ &\quad+\left[1-\operatorname{sgn}[a(z-b)]\right]f(0), \tag*{(19)} \end{aligned}

where f(s)f(s) is any well-behaved test function. An outline of this calculation is given in Appendix.

Gauge transformation function χC\chi_{\mathrm{C}}

The gauge function χC(r,t)\chi_{\mathrm{C}}(\mathbf{r},t) that transforms the Lorentz-gauge potentials ΦL(r,t)\Phi_{\mathrm{L}}(\mathbf{r},t) and AL(r,t)\mathbf{A}_{\mathrm{L}}(\mathbf{r},t) to the Coulomb-gauge potentials ΦC(r,t)\Phi_{\mathrm{C}}(\mathbf{r},t) and AC(r,t)\mathbf{A}_{\mathrm{C}}(\mathbf{r},t) according to

ΦC(r,t)=ΦL(r,t)−∂χC(r,t)c∂t,AC(r,t)=AL(r,t)+∇χC(r,t)(20)\Phi_{\mathrm{C}}(\mathbf{r},t)=\Phi_{\mathrm{L}}(\mathbf{r},t)-\frac{\partial\chi_{\mathrm{C}}(\mathbf{r},t)}{c\partial t},\qquad\mathbf{A}_{\mathrm{C}}(\mathbf{r},t)=\mathbf{A}_{\mathrm{L}}(\mathbf{r},t)+\nabla\chi_{\mathrm{C}}(\mathbf{r},t) \tag*{(20)}

satisfies the Poisson equation

ΔχC(r,t)=∂ΦL(r,t)c∂t,(21)\Delta\chi_{\mathrm{C}}(\mathbf{r},t)=\frac{\partial\Phi_{\mathrm{L}}(\mathbf{r},t)}{c\partial t}, \tag*{(21)}

which follows from the second equality in (20), and the conditions ∂ΦL/c∂t+∇⋅AL=0\partial\Phi_{\mathrm{L}/c\partial t+\nabla\cdot\mathbf{A}_{\mathrm{L}}=0} and ∇⋅AC=0\nabla\cdot\mathbf{A}_{\mathrm{C}}=0 of the Lorenz and Coulomb gauges, respectively.

The Lorenz-gauge scalar potential ΦL(r,t)\Phi_{\mathrm{L}}(\mathbf{r},t) of a point charge qq moving with a constant velocity vz^v\hat{\mathbf{z}} along the zz-axis and passing through the origin r=0\mathbf{r}=0 at a time t=0t=0 is given by

ΦL(r,t)=qs2/γ2+(z−vt)2,γ=11−v2/c2,(22)\Phi_{\mathrm{L}}(\mathbf{r},t)=\frac{q}{\sqrt{s^2/\gamma^2+(z-vt)^2}},\qquad\gamma=\frac{1}{\sqrt{1-v^2/c^2}}, \tag*{(22)}

the partial time derivative of which,

∂ΦL(r,t)c∂t=−qβγ3(z−vt)[s2+γ2(z−vt)2]3/2,β=vc,(23)\frac{\partial\Phi_{\mathrm{L}}(\mathbf{r},t)}{c\partial t}=-\frac{q\beta\gamma^3(z-vt)}{[s^2+\gamma^2(z-vt)^2]^{3/2}},\qquad\beta=\frac{v}{c}, \tag*{(23)}

decays as 1/r21/r^2 at r→∞r\to\infty. When the RHS of Poisson equation (21) is given by (23), the equation cannot be solved using the standard expansion of the inverse distance in spherical coordinates,

1∣r−r′∣=4π∑l=0∞∑m=−ll12l+1r<lr>l+1Ylm∗(θ′,ϕ′)Ylm(θ,ϕ),(24)\frac{1}{\lvert\mathbf{r}-\mathbf{r}'\rvert}=4\pi\sum_{l=0}^{\infty}\sum_{m=-l}^{l}\frac{1}{2l+1}\frac{r_<^l}{r_>^{l+1}}Y_{lm}^{*}(\theta',\phi')Y_{lm}(\theta,\phi), \tag*{(24)}

since the integrand of the requisite radial integral then decays only as 1/r1/r at r→∞r\to\infty, and thus the integral representation of the solution does not converge.

With a positive a=γa=\gamma and b=vtb=vt, and using that 1−γ2=−β2γ21-\gamma^2=-\beta^2\gamma^2, the inverse distributional Laplacian (12) reads

Δˉ−1β2γ3(z−vt)[s2+γ2(z−vt)2]3/2=arsinh⁡z−vts−arsinh⁡γ(z−vt)s,(25)\bar{\Delta}^{-1}\frac{\beta^2\gamma^3(z-vt)}{[s^2+\gamma^2(z-vt)^2]^{3/2}}=\operatorname{arsinh}\frac{z-vt}{s}-\operatorname{arsinh}\frac{\gamma(z-vt)}{s}, \tag*{(25)}

which establishes that

χC(r,t)=qβ[arsinh⁡z−vts−arsinh⁡γ(z−vt)s](26)\chi_{\mathrm{C}}(\mathbf{r},t)=\frac{q}{\beta}\left[\operatorname{arsinh}\frac{z-vt}{s}-\operatorname{arsinh}\frac{\gamma(z-vt)}{s}\right] \tag*{(26)}

is the solution of the distributional Poisson equation

ΔˉχC(r,t)=qβγ3(z−vt)[s2+γ2(z−vt)2]3/2.(27)\bar{\Delta}\chi_{\mathrm{C}}(\mathbf{r},t)=\frac{q\beta\gamma^3(z-vt)}{[s^2+\gamma^2(z-vt)^2]^{3/2}}. \tag*{(27)}

A gauge transformation function equivalent to that of Eq. (26) was obtained in [7]2 by applying a formula of Jackson for χC\chi_{\mathrm{C}} ([8], Eq. (3.6)) to the case of a uniformly moving point charge.

Concluding remarks

The Poisson equation (27) cannot be solved in spherical coordinates, but we solved it using an expansion of the inverse distance in cylindrical coordinates. The choice of the coordinates in which the Laplacian operator is expressed thus may determine whether a given Poisson equation is solvable or not.

Recently, the gauge transformation function χC\chi_{\mathrm{C}} has been found for the case of a point charge set suddenly from rest into uniform motion by integrating with respect to time the first equality in (20), where ΦC\Phi_{\mathrm{C}} and ΦL\Phi_{\mathrm{L}} were the pertinent scalar potentials [9, 10]. The gauge function obtained, which involves arsinh functions (see Eq. (43) in [9]), was confirmed to satisfy the pertinent Poisson equation by evaluating its classical Laplacian. According to the findings of the present paper, when t>0t > 0, the distributional Laplacian adds to that classical Laplacian a term3

qβ[sgn⁡(x)−sgn⁡(x−vt)]δ(s)sΘ(r−ct).(28)\frac{q}{\beta}\left[\operatorname{sgn}(x)-\operatorname{sgn}(x-vt)\right]\frac{\delta(s)}{s}\Theta(r-ct). \tag*{(28)}

It can be shown, however, that the inverse distributional Laplacian of this term vanishes, which makes the extra term (28) in this sense spurious.

Appendix

We outline here an informal proof of the validity of Eq. (19) for any well-behaved test function f(s)f(s). The ‘epsilon-regularized’ Laplacian in (19) evaluates to

Δln⁡[a(z−b)+s2+a2(z−b)2+ϵ2]=(1−a2)A(s2+A2+ϵ2)3/2+ϵ2(B+C),(29)\Delta\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2+\epsilon^2}\right]=\frac{(1-a^2)A}{\left(s^2+A^2+\epsilon^2\right)^{3/2}}+\epsilon^2(B+C), \tag*{(29)}

where

A=a(z−b),(30)A=a(z-b), \tag*{(30)}
B=2(s2+A2+ϵ2)(A+s2+A2+ϵ2)2,(31)B=\frac{2}{\left(s^2+A^2+\epsilon^2\right)\left(A+\sqrt{s^2+A^2+\epsilon^2}\right)^2}, \tag*{(31)}
C=A(s2+A2+ϵ2)3/2(A+s2+A2+ϵ2)2.(32)C=\frac{A}{\left(s^2+A^2+\epsilon^2\right)^{3/2}\left(A+\sqrt{s^2+A^2+\epsilon^2}\right)^2}. \tag*{(32)}

The limit ϵ→0\epsilon\to0 of the first term on the RHS of (A1) is manifestly finite, and so the first term on the RHS of (19) is accounted for immediately. Expanding the test function f(s)f(s) in a Taylor series around s=0s=0, assumed to converge in an interval 0≤s<S0 \le s < S, we can write

∫0∞s ds f(s)ϵ2(B+C)=∑n=0∞f(n)(0)n!ϵ2∫0Sds sn+1(B+C)+ϵ2∫S∞s ds f(s)(B+C).(33)\int_{0}^{\infty} s\,\mathrm{d}s\,f(s)\epsilon^2(B+C)=\sum_{n=0}^{\infty}\frac{f^{(n)}(0)}{n!}\epsilon^2\int_{0}^{S}\mathrm{d}s\,s^{n+1}(B+C)+\epsilon^2\int_{S}^{\infty}s\,\mathrm{d}s\,f(s)(B+C). \tag*{(33)}

Here, the limit ϵ→0\epsilon\to0 of the 2nd term on the RHS vanishes since the integral it involves converges to a finite value when f(s)f(s) is a well-behaved test function.

Let us now investigate the limit

lim⁡ϵ→0[ϵ2∫0Sds sn+1B]=lim⁡ϵ→0[2ϵ2∫0Sds sn+1(s2+A2+ϵ2)(A+s2+A2+ϵ2)2]=lim⁡ϵ→0[2ϵ2∫A2+ϵ2S2+A2+ϵ2du (u2−A2−ϵ2)n/2u(u+A)2].(34)\begin{aligned} \lim_{\epsilon\to0}\left[\epsilon^2\int_0^S \mathrm{d}s\,s^{n+1}B\right] &=\lim_{\epsilon\to0}\left[2\epsilon^2\int_0^S \frac{\mathrm{d}s\,s^{n+1}}{(s^2+A^2+\epsilon^2)\left(A+\sqrt{s^2+A^2+\epsilon^2}\right)^2}\right]\\ &=\lim_{\epsilon\to0}\left[2\epsilon^2\int_{\sqrt{A^2+\epsilon^2}}^{\sqrt{S^2+A^2+\epsilon^2}}\mathrm{d}u\,\frac{\left(u^2-A^2-\epsilon^2\right)^{n/2}}{u(u+A)^2}\right]. \tag*{(34)} \end{aligned}

Here, the integration variable ss was transformed to u=s2+A2+ϵ2u=\sqrt{s^2+A^2+\epsilon^2}, so that ds=u du/u2−A2−ϵ2\mathrm{d}s=u\,\mathrm{d}u/\sqrt{u^2-A^2-\epsilon^2}. For n=0n=0, the RHS of (A6) equals

lim⁡ϵ→02ϵ2[1A(u+A)+1A2ln⁡uu+A]u=A2+ϵ2S2+A2+ϵ2.(35)\lim_{\epsilon\to0}2\epsilon^2\left[\frac{1}{A(u+A)}+\frac{1}{A^2}\ln\frac{u}{u+A}\right]_{u=\sqrt{A^2+\epsilon^2}}^{\sqrt{S^2+A^2+\epsilon^2}}. \tag*{(35)}

This limit vanishes when AA is positive, but when AA is negative, it equals 4, which is contributed by the lower limit of the first term in the brackets,

−lim⁡ϵ→02ϵ2A(A2+ϵ2+A)=−lim⁡ϵ→04A2+ϵ2A=−4∣A∣A=4,(36)-\lim_{\epsilon\to0}\frac{2\epsilon^2}{A\left(\sqrt{A^2+\epsilon^2}+A\right)} =-\lim_{\epsilon\to0}\frac{4\sqrt{A^2+\epsilon^2}}{A} =-4\frac{|A|}{A}=4, \tag*{(36)}

where l’Hopital’s rule is used. For n>0n>0, the integral in (A6) can be evaluated in terms of the Appell function F1F_1 and the hypergeometric function 2F1{}_2F_1. As ϵ→0\epsilon\to0, the arguments of both functions approach ∞\infty; their asymptotic behavior at large arguments can be seen to be as ϵn\epsilon^n ([11]; [12], item 15.7.1), and so their limits ϵ→0\epsilon\to0 vanish.

The limit (A6) is thus

lim⁡ϵ→0[ϵ2∫0Sds sn+1B]=2[1−sgn⁡(A)]δn0.(37)\lim_{\epsilon\to0}\left[\epsilon^2\int_0^S\mathrm{d}s\,s^{n+1}B\right]=2[1-\operatorname{sgn}(A)]\delta_{n0}. \tag*{(37)}

An investigation similar to that in the preceding paragraph yields

lim⁡ϵ→0[ϵ2∫0Sds sn+1C]=−[1−sgn⁡(A)]δn0.(38)\lim_{\epsilon\to0}\left[\epsilon^2\int_0^S\mathrm{d}s\,s^{n+1}C\right]=-[1-\operatorname{sgn}(A)]\delta_{n0}. \tag*{(38)}

Using (A9), (A10) and the other results of this Appendix, we have

lim⁡ϵ→0∫0∞s ds f(s) Δln⁡[a(z−b)+s2+a2(z−b)2+ϵ2]=(1−a2)A∫0∞s ds f(s)(s2+A2)3/2+lim⁡ϵ→0∑n=0∞f(n)(0)n!ϵ2∫0Sds sn+1(B+C)=a(1−a2)(z−b)∫0∞s ds f(s)(s2+a2(z−b)2)3/2+[1−sgn⁡(a(z−b))]f(0),(39)\begin{aligned} \lim_{\epsilon\to0}\int_0^\infty s\,\mathrm{d}s\,f(s)\,\Delta\ln\left[a(z-b)+\sqrt{s^2+a^2(z-b)^2+\epsilon^2}\right] &=(1-a^2)A\int_0^\infty\frac{s\,\mathrm{d}s\,f(s)}{(s^2+A^2)^{3/2}}+\lim_{\epsilon\to0}\sum_{n=0}^\infty\frac{f^{(n)}(0)}{n!}\epsilon^2\int_0^S\mathrm{d}s\,s^{n+1}(B+C)\\ &=a(1-a^2)(z-b)\int_0^\infty\frac{s\,\mathrm{d}s\,f(s)}{\left(s^2+a^2(z-b)^2\right)^{3/2}}+[1-\operatorname{sgn}(a(z-b))]f(0), \tag*{(39)} \end{aligned}

which accomplishes our task.

References

  1. [1]Redžić D V and Hnizdo V 2013 Time-dependent fields of a current-carrying wire Eur. J. Phys. 34 495–501 (arXiv:1301.1573)DOI
  2. [2]Cantelaube Y E 2012 Solutions to the Schrödinger equation, boundary conditions at the origin, and theory of distributions arXiv:1203.0551arxiv.org/abs/1203.0551
  3. [3]Etxebarria J 2023 Acceptable solutions to the radial Schrödinger equation in a central potential Am. J. Phys. 91 792–95DOI
  4. [4]Jackson J D 1999 Classical Electrodynamics (New York: Wiley) 3rd edn
  5. [5]Prudnikov A P, Brychkov Yu A and Marichev O I 1986 Integrals and Series vol 2 Special Functions (Amsterdam: Gordon and Breach)
  6. [6]Vladimirov V S 1979 Generalized Functions in Mathematical Physics (Moscow: Mir)
  7. [7]Hnizdo V 2004 Potentials of a uniformly moving point charge in the Coulomb gauge Eur. J. Phys. 25 351–60 (arXiv:physics/0307124)DOI
  8. [8]Jackson J D 2002 From Lorenz to Coulomb and other explicit gauge transformations Am. J. Phys. 70 917–28 (arXiv:physics/0204034)DOI
  9. [9]Hnizdo V and Vaman G 2024 Potentials and fields of a charge set suddenly from rest into uniform motion Phys. Scr. 99 055534 (arXiv:2311.17652v2)arxiv.org/abs/2311.17652
  10. [10]Hnizdo V and Vaman G 2025 Note on the transformation from the Lorenz gauge to the Coulomb gauge arXiv:2405.16530v4arxiv.org/abs/2405.16530

Paper details

vlnc:W_pvuztdxs
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 3

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents