Alright so here is another thing I do not fully

[cp05]

Alright so here is another thing I do not fully understand.

Inflation is this huge, fast expansion that apparently causes the small perturbations that eventually lead into the structure we see today.

But if I think about it, wouldn't a fast expansion actually eliminate perturbations? Because the universe would be expanding faster than the perturbations can collapse. How is inflation able to create these tiny perturbations instead of destroying them?

 

[phinds]

Hm ... I'm new to this too but my impression is that inflation not only did not create the perturbations, it is what explains why the perturbations are not much larger than they are. Only if inflation happened (well, the only way folks can think of now) is it the case that the CMB varies by only 1/100,000 over the entire observable U.

 

[Chalnoth]

Alright so here is another thing I do not fully understand.

Inflation is this huge, fast expansion that apparently causes the small perturbations that eventually lead into the structure we see today.

But if I think about it, wouldn't a fast expansion actually eliminate perturbations? Because the universe would be expanding faster than the perturbations can collapse. How is inflation able to create these tiny perturbations instead of destroying them?

Well, it tends to do both. Basically, inflation tends to massively dilute anything that happens to be in our universe when inflation begins, so that you end up with a very smooth universe. But there are also zero-point fluctuations of the inflaton field which produce very, very tiny fluctuations.

So it can be sort of thought of as a situation where the inflation cools the universe way, way down, but the quantum fluctuations ensure that a finite temperature remains. This finite temperature, though small, dominates the universe, and leads to the seeds of structure.

 

[Negeng]

I am confused about what caused inflation to begin with. I mean a lot of things I know of like stars and to a lesser degree planets bend the universe "inward" by that I mean the oppisate way inflation would.

 

[Chalnoth]

I am confused about what caused inflation to begin with. I mean a lot of things I know of like stars and to a lesser degree planets bend the universe "inward" by that I mean the oppisate way inflation would.

As of right now, that's an open question. One of the simpler ideas is that inflation itself could be the result of a rare thermal fluctuation in another universe.

 

[bapowell]

I am confused about what caused inflation to begin with. I mean a lot of things I know of like stars and to a lesser degree planets bend the universe "inward" by that I mean the oppisate way inflation would.

Inflation is driven by a special kind of energy. In general, to get inflation going you need to have a sufficiently large region of spacetime in which this kind of energy dominates the total density. This is not a trivial situation to prepare, however. Theories of inflation with chaotic initial conditions, in which the density of the inflation energy is predicted to vary widely across the universe, are somewhat promising.

 

[Tanelorn]

Since it is happening everywhere in free space isn't Inflation just another one of those properties of free space? Only in this case it is a property which has changed with time.
Sorry I keep getting inflation mixed up with dark energy expansion which is what I thought was being discussed.

 

[bapowell]

Alright so here is another thing I do not fully understand.

Inflation is this huge, fast expansion that apparently causes the small perturbations that eventually lead into the structure we see today.

But if I think about it, wouldn't a fast expansion actually eliminate perturbations? Because the universe would be expanding faster than the perturbations can collapse. How is inflation able to create these tiny perturbations instead of destroying them?

As Chalnoth says, inflation indeed smooths things out. Initial perturbations would be wiped clean by a sufficiently long period of inflation. So, classically, inflation renders the universe exceedingly flat and homogeneous. The trick is that the field driving inflation is a quantum field. That means the field has nonzero vacuum fluctuations. One distinctive feature of the inflaton field is that it is a scalar (spin 0) field. This means that it's possible for the field vacuum to have energy. Given that we've just claimed that the field has vacuum fluctuations, what we are ultimately claiming is that the energy of the field fluctuates. And so if you have some region of spacetime that is filled with this field, then even if the energy density is classically uniform, at each point in spacetime you will actually measure some variation in this energy, owing to these fluctuations -- the quantum nature of the field.

As the universe inflates, the classical (average) energy density of the field slowly drops, until it gets to a point at which inflation stops, say $E_{stop}$. Once it stops, the inflaton field turns into ordinary radiation, and the standard cosmology picks up from here. The crucial point here is that, because of these energy fluctuations, different regions of an inflating spacetime will drop to $E_{stop}$ before others -- an initially uniformly inflating patch of spacetime will become speckled with smaller regions in which inflation has stopped. Our observable universe is comprised of a bunch of these regions, all having stopped inflating at slightly different times. The result?? Each of these regions will have a slightly different energy density -- those that ended inflation earlier are underdense relative to those that ended later. These tiny initial inhomogeneities are the seeds of modern day structures. We apparently live in a universe in which all structure -- galaxies, stars, and even people, were born from quantum fluctuations.

 

[Nabeshin]

As the universe inflates, the classical (average) energy density of the field slowly drops, until it gets to a point at which inflation stops, say $E_{stop}$. Once it stops, the inflaton field turns into ordinary radiation, and the standard cosmology picks up from here. The crucial point here is that, because of these energy fluctuations, different regions of an inflating spacetime will drop to $E_{stop}$ before others -- an initially uniformly inflating patch of spacetime will become speckled with smaller regions in which inflation has stopped. Our observable universe is comprised of a bunch of these regions, all having stopped inflating at slightly different times. The result?? Each of these regions will have a slightly different energy density -- those that ended inflation earlier are underdense relative to those that ended later. These tiny initial inhomogeneities are the seeds of modern day structures. We apparently live in a universe in which all structure -- galaxies, stars, and even people, were born from quantum fluctuations.

This is precisely the standard picture of how inflation seeds large scale structure, and if you understand this you're golden. I'd just like to point out, however, that there are several other theoretical mechanisms possibly at play which can also seed structure (I just wrote a paper on this so it's all very fresh in my mind). It's not clear that the fluctuations in the inflation field bapowell mentions are capable of producing the kind of fluctuations we see in the CMB -- in particular, if it turns out that the CMB fluctuations exhibit a significant amount of non-Gaussianity, this is a big problem for this standard approach.

 

[bapowell]

This is precisely the standard picture of how inflation seeds large scale structure, and if you understand this you're golden. I'd just like to point out, however, that there are several other theoretical mechanisms possibly at play which can also seed structure (I just wrote a paper on this so it's all very fresh in my mind). It's not clear that the fluctuations in the inflation field bapowell mentions are capable of producing the kind of fluctuations we see in the CMB -- in particular, if it turns out that the CMB fluctuations exhibit a significant amount of non-Gaussianity, this is a big problem for this standard approach.

Good point. I'm guessing you have curvatons on the brain? Although depending on the type of NGs, this could still be the correct picture.

 

[Chalnoth]

This is precisely the standard picture of how inflation seeds large scale structure, and if you understand this you're golden. I'd just like to point out, however, that there are several other theoretical mechanisms possibly at play which can also seed structure (I just wrote a paper on this so it's all very fresh in my mind). It's not clear that the fluctuations in the inflation field bapowell mentions are capable of producing the kind of fluctuations we see in the CMB -- in particular, if it turns out that the CMB fluctuations exhibit a significant amount of non-Gaussianity, this is a big problem for this standard approach.

Slightly technical point...but about what degree of $f_{NL}$ would be significant in this context?

 

[Nabeshin]

Slightly technical point...but about what degree of $f_{NL}$ would be significant in this context?

I don't have the citation off hand but I remember reading standard slow-roll inflation is only capable of producing $f_{NL} \lesssim 1$. And there is that paper which claims to find $27 < f_{NL} < 147$. [1]

bapowell: Yes, curvatons are one of the things on my mind :) The other main one is the modulated reheating scenarios, but now we're getting off topic.

[1]: http://arxiv.org/abs/0712.1148

 

[Chalnoth]

I don't have the citation off hand but I remember reading standard slow-roll inflation is only capable of producing $f_{NL} \lesssim 1$. And there is that paper which claims to find $27 < f_{NL} < 147$. [1]

Well, when the error bars are that large, they're really consistent with zero. It is hoped that Planck will be able to detect an $f_{NL}$ with an error of about $\pm 5$. So in any case it should be clear when the Planck CMB results are released whether or not this is a detection.

 

[bapowell]

The range given by Nabeshin I believe is 95% CL, so those are the error bars. If that particular measurement, which is based on LSS, holds, then it will give a nonzero detection of non-Gaussianity at 2 sigma. Indeed, Planck should get fnl down to 5 or so for local-type NGs, but not do as well for equilateral (for which it might achieve fnl \sim 25). But, yes, we are getting off track.

 

[Chalnoth]

The range given by Nabeshin I believe is 95% CL,

Yes, 2-3 sigma is extremely common for spurious detections. So yeah, consistent with zero so far.