March 14, 2005

Einstein Model [Open Thread]

Posted by Arcane Gazebo at March 14, 2005 2:05 PM

Happy Pi Day, and happy 126th birthday to Albert Einstein.

2005 is the 100th anniversary of Einstein's "miracle year"; therefore, during that year Einstein was as old as I am now. There's a humbling thought.

Cass McCombs: Prefection: The first time I listened to this I was driving on the 880 with the windows down, so I mainly heard the percussion and (barely) the vocals (and the occasional organ), which made it sound more hard-rock than it actually is. Later on headphones I realized that there were lots of broadband* elements that I didn't appreciate before. It's interestingly textured. Inexplicably, after the last track plays there's a recording of a car alarm with sounds of traffic in the background. I kept listening to see if anything would happen... for seven minutes. After which the alarm fades out to a radio announcer saying, "Up next, a new treatment for obsessive-compulsive disorder." Yeah, very funny. For your listening pleasure, here's "Subtraction".

*This is the danger of allowing physicists to write about music.


In the nonlinear dynamics seminar I run, we had a seminar on predator-prey models today. I was hoping to schedule that talk for Valentine's Day, but things didn't work out...

Posted by: Mason | March 14, 2005 8:06 PM

Yay for open threads!

I just got done reading Market Forces, by Richard K. Morgan (Author of Altered Carbon and Broken Angels). You'd think being done with classes for the term would inspire games-playing, but i'd just gotten that book (as well as Shadowmarch, by Tad Williams). Anyways, my review:

It had all the things I liked about Morgan's previous books except for one, which was a main character I liked in the end. I really was quite dubious about the main character and still don't know what to think of him. Things I liked about his previous books that apply to this one: Absolutely stunning world setting. Just one of those vividly imagined alternate worlds (not the same one as Altered Carbon) that is great to read stories in, just to see more of the world and how it works. Very atmospheric and very well-drawn characters.

As far as the plot goes, it was well done and hard to tell where it was going at any particular point in it. The ending was completely not what I wanted to happen, something just didn't click with it and the previous parts of the book. Still, a good read, especially if you liked Morgan's previous works. :)

Ok... late night open thread posting done! Rate me on how incoherent it was. :)

Posted by: Zifnab | March 17, 2005 5:31 AM

Oh wow, I really should read that. After I give this talk next week I'll actually have time to read, too.

I still have a big stack of books waiting for me, but I tend to give priority to newer ones. So after I finish Smilla's Sense of Snow it'll probably be Market Forces and Kafka on the Shore (not necessarily in that order).

Posted by: Arcane Gazebo | March 17, 2005 10:26 AM

Hey, don't mean to hijack the topic, but after all, this is an open thread...

I've got a question, mainly for the Gazebo, but for anyone who is willing to give an opinion as well.

Say I have a solid, perfectly conducting sphere. Say that I have an oscillating magnetic dipole somewhere inside this sphere. Is the total magnetic field outside of the sphere zero?

I was typing up an alternate problem as well, but it just turns into a mess when I try and word it, so I'll stick with the one above.

Posted by: Lemming | March 17, 2005 1:33 PM

Lemming: The answer seems to be yes. There's probably a quick way to get there from Maxwell's equations, but I consulted Jackson who does a more general case in section 5.18. He assumes quasi-static fields, a single conductor with conductivity \sigma and permeability \mu, and finds that the vector potential A obeys a diffusion equation:

\nabla^2 A = \mu \sigma \partial A / \partial t

For a perfect conductor \sigma goes to infinity, so we need \partial A / \partial t = 0, so the overall magnetic field is constant in time. It seems to me that any constant magnetic field would have to be proportional to the time average of the dipole moment, so if the dipole is oscillating in a symmetric fashion this would also be zero.

All this is for inside the conductor, but for the sphere with dipole inside, if B=0 inside the sphere it is certainly also 0 outside.

Posted by: Arcane Gazebo | March 17, 2005 2:06 PM

If you could justify your final assertion, I'd really appreciate it--that one strikes me as odd, not in an intelligent but in more of a pattern matching sort of way, because the surface integral equations we use (if you know the value of the electic and magnetic fields on a surface that result from a source on one side of the surface, you can compute the value of the resultant field anywhere on the *other* side, or on, the surface) result in a sensible field on one side of the surface, and go identically to zero on the other.

That's sort of hand-wavy and, like I said, pattern matching more than anything else. I just write the damn code.

Posted by: Anonymous | March 17, 2005 3:26 PM

And I've also just been corrected--I'm apparently talking about a *hollow* sphere, not a solid one.

Sorry. Things stay abstract so often that I completely lose track of the physical technicalities of the system we're talking about. >

Posted by: Lemming | March 17, 2005 4:04 PM

I was going to say something about how B=0 on the outer surface of the sphere gives you a boundary value problem where B=0 everywhere is a solution. But now I'm doubting my answer, because certainly if it were an electric field there would be surface charges on the sphere even though the field inside is 0. I should spend a few more minutes thinking about this...

Posted by: Arcane Gazebo | March 17, 2005 4:15 PM

So I guess the question is, can the dipole field be completely canceled by currents on the interior surface? Or does one need currents on the exterior surface as well, which would create a dipole field outside the sphere? I feel like the second case is a way of cutting magnetic field lines, which I seem to recall isn't possible. But maybe that's not the right way to think about it.

Posted by: Arcane Gazebo | March 17, 2005 4:32 PM

I'm still not 100% positive whether the problem I'm tackling is a hollow infinitely-thin shell, or a solid sphere.

I can tell you this, though--I can independently compute a set of surface currents on the outter surface of the sphere that, when you treat them as a bunch of electric dipoles and integrate, I can solve to make them match (outside) with an arbitrary dipole (inside). Only in the case of the dipole being at the origin, however, do my solved surface currents actually run tangent to the surface.

I don't know how helpful that is, but that's the code I got working today, and it was, IMHO, a *quite* clever engineering/programming feat (mathematics aside) to trick our code into computing it (a substantially different computation than we normally do) by only changing a few lines of code. At least it's *one* thing for me to feel good about.

Maybe I'll just harass you when you're down here later this month.

Posted by: Anonymous | March 17, 2005 4:58 PM

Greetings from the Las Vegas airport, where I am stranded. (I'm taking my passing similarity to Tom Hanks to new extremes with my impersonation of his role in Terminal.)

I am going to be quite late for my interview today, but at least I have a flight that will get me to Fresno at 10:26 (barring any other unforseen disasters, of which there have been several lately when it comes to flights).

I hope they're forgiving. By the way, there's a chance I'll fly back up there on 3/24 or 3/25 and forgo post-talk parts of the March Meeting and (sadly) D & D as well. It may be necessary for me to do this to still have a chance at this job, as I am going to miss half my interviews today. I also hope that my talk and interviews are somewhat coherent given that by the end of the day I will have been awake over 40 hours in a row.

Sigh...this was not how I envisioned things playing out...

As if my travel schedule weren't hectic enough. There are many good things happening (in theory) with all of these trips, but some of them were scheduled before I got the 'come hither right away' messages of recent weeks. So I look forward to all this stuff with a mixture of excitement and dread. (Excitement for the cool stuff---like a free trip to Sydney and seeing friends in LA---and dread for long hours in planes and airports.

For my ranting pleasure (and your pain in perhaps my version of misery loves company):

3/17: Atlanta-->Vegas-->Fresno [alteration of an Atl-->LA flight, which caused other problems not in the schedule per se]

* now to be finished today 3/18

3/19: Fresno--> LAX (and then ignore the last part of the flight that goes to Orange County)

[work on March Meeting talk, although I might still do some of that this morning; the problem is I'm far more in the mood for ranting and reading]

3/27: LAX--> Atl (but not the original flight I had, because those bastards have an obnoxious policy that I'll explain in another rant; it's one of those things that is technically legal but logically makes one go wtf)

3/28: Atl--> Toronto (visit collaborator at McMaster)

3/30: Toronto-->Atl

around 4/3: Atl--> Sydney (24 hours in the plane, basically)

around 4/9: Sydney--->Atlanta

**stay in Atlanta for a week**

around 4/17: Atl--> Berkeley

around 4/23: Berkeley--> Atlanta

**stay in Atlanta for about 3 weeks and then go to a conference in Taiwan, but by then I should have recovered**

I know I should just do work instead of rant, but some things never change. I need a bitch session, and online approximations of it are just going to have to do.

I look forward to hopefully not traveling too much this summer. I'm already exhausted.

Posted by: Mason | March 18, 2005 6:29 AM

Amusing typo for the day:

Meant to type, "takes a long time." (Regarding a piece of code.)

Ended up typing, "takes a logn time."

Posted by: Lemming | March 18, 2005 2:23 PM

Mason: Wow, that's terrible. If you have to leave the meeting early we'll have to have some sort of games night before then. :)

Posted by: Arcane Gazebo | March 19, 2005 3:00 PM

The people at Merced were nice and are not counting this against me. (There may or me not be a couple extra phone conversations, but they were quite accomodating.)

By the way, I don't recommend going to a job interview after an all-nighter. The adrenalin (and the sugar and caffeine) kept me going, and the only visible cue to my tired state was the extra shaking in the laser pointer from all the caffeine. By the time I crashed, I had been up for 40 hours consecutively, which is not new but definitely not optimal for an interview. (I kept talking to people because any idle time would have caused me to nod off.)

Anyway, I am one of 4 interviewing for applied math and am also being considered for their physics spot, so hopefully it will work out favorably... designing a program from scratch would be way cool...

Posted by: Mason | March 25, 2005 9:24 AM
Post a comment