On hadron colliders, dark matter and black holes

[Karro]

I just read this in the news a minute ago. Naturally my first reaction was to come here. Actually, naturally, my first reaction was "WTF?"... then to come here and ask you guys "WTF?". That's nuts.

Maybe this guy bought into the various sci fi scenarios that've been hashed out, here and in other places. I guess that would make him a poor scientist/engineer...?

[torstan]

Dan Brown has even more to answer for than we had first assumed. Perhaps it was all a plan to make CERN a honey trap for deluded terrorists...

It actually seems more prosaic than that. The guy is an engineer for an affiliated organisation, rather than CERN itself and he was apparently involved in suggesting French targets. So not really anything to do with CERN, but I guess we'll have to wait to find out any more. Certainly no suggestion that I've seen that he was trying to sabotage/steal/misuse anything being done at CERN.

In other news - the injection set-up that gets the protons up to the speed they need for the injection into the LHC has been tested and is working fine. All seems to be going smoothly.

[torstan]

The LHC is now cold. All sectors have been cooled down to their operating temperature. So fingers crossed, all is going well and the final stages of commissioning are underway.

[RPMiller]

Torstan, are you involved with this project?

http://www.youtube.com/watch?v=gCgTJ...eature=related

[torstan]

I'm not, but it certainly has an impact on my work. The mapping of large scale structure is crucial to figuring out the distribution of dark matter. Once we have that nailed down (and its getting better all the time) we can try to recreate it in simulations. So what you do is create a vast numerical simulation of a load of fairly evenly distributed dark matter in the very early universe and then you let it evolve. The gravitiational attraction of a small clump of dark matter pulls in more dark matter over time until the small clumps become large clumps. These orbit even larger clumps and so the distribution goes from smooth to very lumpy. Now these lumps have a large gravitational pull, and the normal matter gets pulled into the middle of them. This is how galaxies form.

We've got a pretty good idea of just how smooth the distribution of the dark matter was in the early universe and we now know (thanks to surveys like the one you linked to) how the dark matter is distributed today. So the numerical simulations can vary the properties of their dark matter model to get the simulation to match up with those boundary conditions. This helps us to pin down the properties of the dark matter.

My interaction with this is pretty much to keep an eye on the limits they come up with and make sure that any model I work on doesn't break those simulations.

[torstan]

Oh, and here's a link to quite a nice video of the Millenium Simulation, one of these numerical simulations and one of the largest ever done. Think of it as a map of all the matter in a possible alternate universe. The challenge is to get the rough features looking about right - a little like creating a map with model of erosion and trying to make sure that none of the rivers violate the river police.

I'm not sure the music adds much though so you might want to turn it off before running through it.

[Redrobes]

Your model, is it a pure math model or a software simulation ? If like a simulation then like GTS or perhaps a weather prediction sim these are not exact and open to a lot of parameter twiddling. Is the idea to have a feel for what parameters might give the right output. If math then is the math so complex that its not possible to come up with a proof. E.g. Ye olde E=mc^2 was a theoretical proof that needed experimental evidence but some of the nuke simulations done on the high end iron or the fluid dynamics and finite element analysis is still just numerical modeling.

Here's another question(s). Whats the simplest pure math prediction (i.e. as yet unknown physics) that has been made that the LHC might be able to show evidence for and maybe a curve ball, whats the weirdest and most fantastic claim that any scientist has made that there is a real possibility that the LHC might with a small chance actually show.

[Steel General]

Interesting...and I didn't mind the music (but I am a Pink Floyd fan )

[torstan]

Simulations of large complex systems are always numerical simulations. This is true for the fluid dynamics in the simulation of nuclear explosions and it's certainly true when you're trying to build a universe. However, the physics underlying the numerical simulation is good. We have solid models of a basic dark matter particle - it's massive and it interacts through gravity, but it doesn't interact strongly in any other way. This means you just set up a simulation with the laws of gravity built in and a number of particles that those laws affect and let it run. The more powerful the computer, the more particles you can place in the simulation.

At the moment each particle is actually an enormous blob of dark matter - with a mass many times that of our sun (roughly 10^4 solar masses if you're simulating a galaxy, right up to 10^12 solar masses if you're looking at recreating the whole observable universe). These blobs interact through gravity and form the clumps and stringy filaments that you see in the millenium simulation. As super computers get better, the mass of each clump is reduced (and the number of clumps consequently increased so the total mass is the same) to see if the more fine grained interactions change anything.

So within the framework of the model - we know the physics exactly. Solving that for billions of particles is impossible, so we run the simulations. If we find some features in our surveys that the simulation can't explain, then we will see if we need to change the physics model (by adding in some interactions between dark matter particles for example) or whether there's some problem with the numerical simulation (perhaps that its not fine grained enough).

In the case of the second question - the simplest pure maths prediction at the LHC is the Higgs boson. It has to be there for the symmetry of the theory to work, and for the amplitude of the scattering of the W bosons not to go to infinity as the energy increases. So that's a pretty clear pure maths prediction. I guess the simplest version of that is that there has to be new physics in the energy range to stop the probability of a specific result in WW scattering becoming greater than 1 (that's generally considered a bad thing). Even if there's no Higgs boson, we know that something new must happen in the energy regime of the LHC. So no matter what, it will find some deviation from physics as we know it.

As for the weird stuff? Well there is a lot of that. Nielsen recently claimed that the universe would conspire against creating the Higgs boson so that CERN would never start. I guess that would be a dramatic example - but I think that's pretty tongue in cheek so shouldn't be taken seriously. Large extra dimensions is probably one of the strangest things that might show up. We might probe down to a length scale where we would be able to see the effect of extra dimensions - and some of the created particles would travel in them. Equally, we might produce microscopic black holes that would decay in a burst of Hawking radiation. That would be really cool. I think those are some of the more bizarre things that have been proposed.

The other thing that we are always told by those scientists that have been through this process before is that experiments always end up showing things that you never predicted. Nature tends to be more inventive than physicists, so we may see something truly bizarre.

[Redrobes]

Cool - I did hear about that universe trying to prevent the LHC... worth a small chuckle.

Measureable - as in not requiring a ruler the size of Geneva - extra dimensional spaces would be cool and if you can see what passes through them then I guess thats part way there to knowing how or what to do to build them bigger.

Y'know I am still impressed when you walk down a corridor with one of those little boxes thats like the PIR detectors but have the little radars in them. At the middle of the radar is a little diode which makes the radar. Its a tunnel diode and its a quantum device. I have one here abouts somewhere. A real quantum device. What, 40 odd years ago noone knew anything about quantum stuff and now I have these on a shelf. Nobody knows how these LHC discoveries will filter down to consumer items. Dibs on the first portable hole tho...

The Higgs must be really exciting then to know that theres going to be some new physics cos it just cant not be there otherwise the maths fails. Thats the sort of thing thats cool as you know what your looking for but not what it is. Its not like not knowing what your looking for. Thats like code debugging. You just slice and dice until you find what the issue is. Instead of a typo cut n paste error tho, its a new branch of physics that pops out. Thats pretty cool.

The sim thing sounds a little more dodgy. I have GTS and it has the "water flows downhill" bit of hard physics in it and its always producing terrain thats not quite right and there's always more stuff to add in to make it closer to the desired observable result. But if I get good results from it I don't think its true that the model is right. For a start I guessed all of the math in mine. I didn't even so much as look at any previous works or papers. So its all wrong but looks about right give or take. I could use the app to predict stuff but I still don't think that it makes the underlying model right.

How do you think you know that a sim model is right or is it that you never know but it might allow you to run interesting tests to exercise sim cases to narrow down the math / physics tests required in real life or on the LHC ? Is it like, the sim says at 1.234 TeV you get a spike so lets wind it up and see. If yea then sim looks like it could be right, if nae then its back to the drawing board.

On a side note too. The Hawking radiation. I read that the smaller the black hole the more of it you have which makes small holes unstable. Is that from area of sphere / vol of sphere goes up as r -> 0... anyway, I take it that the radiation comes off as photons and I guess that its not necessarily emitted uniformly - or is there a lot of it so much that it makes no odds. But my question was, How do you know the light came from a black hole instead of something else. I thought that all collisions and electron decay chucked out photons or stuff that turns into photons and I thought that you were making a lot of collisions at this point. Is that all done in the post processing ?

The whole way this thing works is like a wall of fog to me.