09.01.2014 Views

Pictures Paths Particles Processes

Pictures Paths Particles Processes

Pictures Paths Particles Processes

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

84 November 7, 2013<br />

k<br />

↔<br />

¯h<br />

| ⃗ k| 2 + m 2<br />

k 1 k 2<br />

↔ − λ 4<br />

k 4 k 3<br />

¯h (2π)D δ D ( ⃗ k 1 + ⃗ k 2 + ⃗ k 3 + ⃗ k 4 )<br />

k 1 k 2 ↔ +<br />

J( ⃗ k 2 )<br />

¯h (2π)D δ D ( ⃗ k 1 + ⃗ k 2 )<br />

At vertices, the wavevectors are considered either all incoming or<br />

all outgoing.<br />

Each internal wave vector ⃗ k is to be integrated over, with integration<br />

element d D⃗ k/(2π) D .<br />

Feynman rules, version 2.5 (2.85)<br />

2.4.6 Loop integrals<br />

As stated above, diagrams with loops contain internal wave vectors that have<br />

to be integrated over, and many of these integrals are divergent. Therefore, we<br />

have two face two technical challenges. In the first place, we have to devise a<br />

way to quantify these divergences : this is called regularization. In the second<br />

place, regularizing these divergences does not make them go away, and therefore<br />

we shall have to arrive at a method of including these divergences into the<br />

theory in such a way as to yield finite and unambiguous answers for physically<br />

interesting quantities. This last procedure is called renormalization. In this<br />

section we shall only address regularization, for the case of one-loop integrals.<br />

The idea of regularization is to let the theory depend on an arbitrarily introduced<br />

parameter, such that the divergences appear when that parameter takes<br />

on a certain value. Different regularization schemes are available, with different<br />

choices for the extra parameter, which may be particle masses, upper limits on<br />

momenta, etcetera. It must be kept in mind, however, that theories may depend<br />

sensitively on such parameters, and therefore it may be prudent to choose the<br />

parameter in such a way that the behaviour of the theory does not depend on it<br />

too sensitively. The most popular regularization scheme is that of dimensional<br />

regularization : in this approach the number of dimensions, D, is chosen as the<br />

freely varying parameter. Already anticipating that we shall study theories in<br />

four spacetime dimensions, we therefore write<br />

D = 4 − 2ɛ ,<br />

with the implication that, at the end of all calculations, we shall take ɛ down to

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!