量子电动力学中含Heaviside函数的积分求解问询

Hey, let's break down this integral step by step—since you're working in QED, these Fourier-style integrals with Heaviside functions rely on standard tools from distribution theory and contour integration, which are bread-and-butter here.

First, let's recap your rewritten integral to make things clear:
$$
I = \int_{-\infty}^{+\infty} k^2 \mathcal{H}(k) e^{ick\left(t'-t+\frac{r-R}{c}\right)}dk -\int_{-\infty}^{+\infty} k^2 \mathcal{H}(k) e^{ick\left(t'-t-\frac{r-R}{c}\right)}dk = I_1 - I_2
$$

Step 1: Simplify the Integral Range with $\mathcal{H}(k)$ #

The Heaviside function $\mathcal{H}(k)$ is 1 for $k>0$ and 0 for $k<0$, so we can immediately restrict both integrals to $k \in [0, \infty)$:
$$
I_1 = \int_{0}^{\infty} k^2 e^{ick\tau_1}dk, \quad I_2 = \int_{0}^{\infty} k^2 e^{ick\tau_2}dk
$$
where $\tau_1 = t'-t + \frac{r-R}{c}$ and $\tau_2 = t'-t - \frac{r-R}{c}$. These integrals are formally divergent, so we'll handle them in the distributional sense (standard for QED calculations).

Step 2: Regularize the Integrals with the +iε Prescription #

To make the integrals converge, we introduce a small positive imaginary part $\epsilon$ to the exponent (the Feynman+iε prescription), then take $\epsilon \to 0^+$:
$$
I_1^\epsilon = \int_{0}^{\infty} k^2 e^{ick\tau_1 - \epsilon k}dk, \quad I_2^\epsilon = \int_{0}^{\infty} k^2 e^{ick\tau_2 - \epsilon k}dk
$$
We can compute these using the standard integral formula $\int_{0}^{\infty} k^n e^{-bk}dk = \frac{n!}{b^{n+1}}$ (valid for $\text{Re}(b) > 0$). For $I_1^\epsilon$, set $b = -ic\tau_1 + \epsilon$:
$$
I_1^\epsilon = \frac{2!}{(-ic\tau_1 + \epsilon)^3} = \frac{2}{(-i)^3(c\tau_1 - i\epsilon)^3} = \frac{2i}{(c\tau_1 - i\epsilon)^3}
$$
The expression for $I_2^\epsilon$ is identical, just replace $\tau_1$ with $\tau_2$.

Step 3: Take the Distributional Limit ($\epsilon \to 0^+$) #

Now we need to interpret $\frac{1}{(x - i\epsilon)^3}$ as $\epsilon \to 0^+$. Using properties of distributions:
$$
\lim_{\epsilon \to 0^+} \frac{1}{(x - i\epsilon)^3} = \text{P.V.}\left(\frac{1}{x^3}\right) + i\pi \delta''(x)
$$
where $\text{P.V.}$ denotes the Cauchy principal value, and $\delta''(x)$ is the second derivative of the Dirac delta function. Substituting back into $I_1$ and $I_2$:
$$
I_1 = \frac{2i}{c^3}\left( \text{P.V.}\left(\frac{1}{\tau_1^3}\right) + i\pi \delta''(\tau_1) \right) = \frac{2i}{c^{3}\text{P.V.}\left(\frac{1}{\tau_1}3}\right) - \frac{2\pi}{c^3}\delta''(\tau_1)
$$
$$
I_2 = \frac{2i}{c^{3}\text{P.V.}\left(\frac{1}{\tau_2}3}\right) - \frac{2\pi}{c^3}\delta''(\tau_2)
$$

Step 4: Combine Results and Apply Physical Constraints #

Subtract $I_2$ from $I_1$:
$$
I = \frac{2i}{c^3}\text{P.V.}\left( \frac{1}{\tau_1^3} - \frac{1}{\tau_2^3} \right) - \frac{2\pi}{c^3}\left( \delta''(\tau_1) - \delta''(\tau_2) \right)
$$
Now, recall your original integral is over $t' \in [0, t]$. Let's analyze the delta function terms:

• $\delta''(\tau_2) = \delta''\left(t' - \left(t + \frac{r-R}{c}\right)\right)$: Since $t + \frac{r-R}{c} > t$ (assuming $r > R$), this delta function has no support in $[0, t]$ and can be ignored.
• $\delta''(\tau_1) = \delta''\left(t' - \left(t - \frac{r-R}{c}\right)\right)$: This only contributes if $t - \frac{r-R}{c} \geq 0$ (i.e., enough time has passed for the signal to propagate from $R$ to $r$).

For the principal value term, we can simplify the algebraic expression:
$$
\frac{1}{\tau_1^3} - \frac{1}{\tau_2^3} = \frac{\tau_2^3 - \tau_1^{3}{(\tau_1\tau_2)}3} = \frac{-2\frac{r-R}{c}\left(3(t'-t)^2 + \left(\frac{r-R}{c}\right)^2\right)}{\left( (t'-t)^2 - \left(\frac{r-R}{c}\right)^2 \right)^3}
$$
Substituting back gives the explicit form of the principal value contribution.

Step 5: Physical Interpretation #

In QED, this integral describes the interaction of the conjugate momentum field $\hat{\pi}(t')$ with the propagator (Green's function) for the field. The delta function term corresponds to instantaneous contact interactions, while the principal value term describes the long-range, retarded/advanced field interactions—consistent with the causal structure of QED.

内容的提问来源于stack exchange，提问作者Arturo Villegas