07.01.2014 Views

PHYS08200605006 D.K. Hazra - Homi Bhabha National Institute

PHYS08200605006 D.K. Hazra - Homi Bhabha National Institute

PHYS08200605006 D.K. Hazra - Homi Bhabha National Institute

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

4.3. THE NUMERICAL COMPUTATION OF THE SCALAR BI-SPECTRUM<br />

In the equilateral limit,<br />

f NL<br />

(k 1 ,k 2 ,k 3 ) simplifies to<br />

the expression (1.29) for the non-Gaussianity parameter<br />

f eq<br />

NL (k) = −10 9<br />

1<br />

(2π) 4 k 6 G eq (k)<br />

P 2 S (k) , (4.34)<br />

where P S<br />

(k) is the scalar power spectrum defined in Eq. (1.21). It is straightforward to<br />

show that the f NL<br />

corresponding to the super-Hubble contribution to the bi-spectrum<br />

G es<br />

eq (k) above is given by<br />

f eq(se) (k) ≃ − 5iM2 Pl<br />

NL<br />

18<br />

( ) [ A<br />

2<br />

k ¯B∗ 2 k −A ∗ k 2 ¯B2 k<br />

12I<br />

|A k | 2 13 (η e ,η s )− 9 ]<br />

4 I 56(η e ,η s ) , (4.35)<br />

where we have made use of the fact that f k ≃ A k at late times in order to arrive at the<br />

power spectrum.<br />

To estimate the above super-Hubble contribution to the non-Gaussianity parameter<br />

f eq(se) , let us choose to work with power law inflation because it permits exact calculations,<br />

and it can also mimic slow roll inflation. During power law expansion, the<br />

NL<br />

scale<br />

factor can be written as<br />

a(η) = a 1<br />

( η<br />

η 1<br />

) γ+1<br />

, (4.36)<br />

where a 1 and η 1 are constants, while γ is a free index. It is useful to note that, in such a<br />

case, the first slow roll parameter is a constant and is given by ǫ 1 = (γ + 2)/(γ + 1). The<br />

current observational constraints on the scalar spectral index suggest that γ −2, which<br />

implies that the corresponding scale factor is close to that of de Sitter.<br />

In power law inflation, the exact solution to Eq. (1.16) can be expressed in terms of the<br />

Bessel function J ν (x) as follows (see, for instance, Refs. [105]):<br />

v k (η) = √ −kη [C k J ν (−kη)+D k J −ν (−kη)], (4.37)<br />

where ν = (γ +1/2), and the quantities C k and D k are constants that are determined by<br />

the initial conditions. Upon demanding that the above solution satisfies the Bunch-Davies<br />

initial condition (1.23), one obtains that<br />

C k = −D k e −iπ(γ+1/2) ,<br />

(4.38a)<br />

√ π e iπγ/2<br />

D k =<br />

k 2 cos(πγ) . (4.38b)<br />

Sincef k = v k /z, withz = √ 2ǫ 1 M Pl<br />

a, and asǫ 1 is a constant in power law inflation, we can<br />

arrive at the constants A k and B k [cf. Eqs. (4.14) and Eqs. (4.15)] from the super-Hubble<br />

71

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

Saved successfully!

Ooh no, something went wrong!