0
$\begingroup$

I am reading Peskin and Schroeder's chapter on functional methods. They propose the amplitude:

$$ \langle \phi_b(\vec{x})|e^{-iHT}| \phi_a(\vec{x})\rangle = \int \mathcal{D}\phi\mathcal{D}\pi \exp \bigg[ i\int_0^T \left( \pi \dot\phi - \frac{1}{2}\pi^2 - \frac{1}{2} (\nabla\phi)^2-V(\phi) \right) \bigg] $$ They then said that we can complete the square and integrate over $\mathcal{D}\pi$ integral. How is this done?

I was able to complete the square: $\pi\dot\phi - \frac{1}{2} \pi^2=-\frac{1}{2}(\pi-\dot\phi)^2+\frac{1}{2}\dot\phi^2$ and rewrite the amplitude as:
$$ \langle \phi_b(\vec{x})|e^{-iHT}| \phi_a(\vec{x})\rangle = \int \mathcal{D}\phi \exp \bigg[ i\int_0^T \left( \frac{1}{2} \partial_\mu\phi\partial^\mu\phi-V(\phi) \right) \bigg] \int \mathcal{D}\pi \bigg[ -\frac{1}{2}i \int^T_0 (\pi-\dot\phi^2) \bigg] $$ I am only confused about the functional integral (mathematically), how is it done?

$\endgroup$
2
$\begingroup$

You do the following

$$\int D\pi \exp[ -\frac12 i \int_0^T (\pi-\dot \phi)^2] =\int D\pi \exp[-\frac12 i\int_0^T \pi^2]= constant$$

where I've shifted $\pi\to \pi+\dot \phi$ which leaves the measure $D\pi$ invariant. The result is a $\phi$ independent multiplicative constant. It's not quite well-defined; you'll need to determine this constant by other means, if necessary.

$\endgroup$
0
$\begingroup$

well functional integral can be thought as vector integral

$$I=\int d\vec{v}e^{-\frac{1}{2}\vec{v}^tA\vec{v}}=\prod_{a}\int dv_ae^{-1/2u_a^2(\lambda_{a})}=\sqrt{\det(2\pi A^{-1})}$$

where $\lambda_a(u_a)$ is eigenvalues(vectors) of $A$ the above calculation is very easy gaussian vector integral. Now instead of discrete vector we have continuous "vectors"

$$I=\int dv(x)e^{-\frac{1}{2}\int dx\int dy\ v(x)A(x,y)v(y)}=\prod_{x}\int du(x)e^{-1/2u(x)^2(\lambda(x))}\propto\sqrt{\det( A^{-1})}$$

in your case,

$$I=\int dv(x)e^{-\frac{1}{2}\int dx\int dy\ v(x)A(x,y)v(y)+\int dxj(x)v(x)}\propto\sqrt{\det( A^{-1})}\int d v(x)e^{\frac{1}{2}\int d x\int dy v(x)A^{-1}(x,y)v(y)}$$ so this last equation is automatically square completed version, just apply $\pi$ integral to your case in your case: $A=i$ and$j=i\dot{\theta}$ so if we apply our formula we get, $$\text{constant}\times e^{i1/2\int dx\dot{\theta}^2}$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.