|
|
|
|
|
|
Posted: Sat Aug 04, 2007 8:51 pm
I understand, generally and vaugely, how the uncertanty principle works, but not on a technical or conceptual level. Can someone help me with some of that?
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Sun Aug 05, 2007 12:08 am
How much mathematics do you know? Because formally the uncertainty principle requires some calculus and some abstract algebra and a bit of statistics.
Conceptually, it works as following:
When you're not observing it, a particle exists in what is called a "superposition" of states, meaning that the only way to describe the particle's position is to say that it might be in any position (with varying, determinable probabilities) and the only way to describe the particle's momentum is to say that it might have any momentum (once again, with varying, determinable probabilities). When you observe that particle's position, you force the particle into a position state, which has only one value for the position. Similarly, when you observe the particle's momentum, you force the particle into a momentum state, which has only one value for the momentum. But, sadly, when you push the particle into a position state, you also push it into a momentum superposition, so now it might have any momentum. Similarly, when you push the particle into a momentum state, you also push it into a position superposition (terrible terminology), so now it might have any position. Because superpositions can only be described probabilistically, we have some uncertainty as to the value the particle has for the quantity for which it's in a superposition. Since there this incompatibility, we get that the uncertainty in position due to superposition times the uncertainty in momentum due to superposition cannot be less than a certain number, which then forces the particle to enter a superposition of one measurement every time it leaves a superposition of the other.
Now, of course, we come to the question of "so what is the momentum of a particle in a position state?" The answer is "we don't know, we don't know if a particle in a position state has a meaningful momentum, and we don't know if particles actually have a physically meaningful existence."
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Sun Aug 05, 2007 8:16 am
*collaborates with a small part of the math behind it >_>*
Basically, you have two amounts of standard deviations. This in statistics means how spread out the extreme values of a measurement are. The more standard deviations we have, statistics tell us we have less certainty of something. In here, we use the letter delta (Δ) to represent this. We have the deltas of position, and momentum (Δx and Δp). The uncertainty principle says that the multiplication of these values has to be greater than a constant, which basically means that no matter how hard you try, you won't get more precision in both things at the same time. The mathematical notation of this is
ΔxΔp ≥ ħ/2
Where ħ (h-bar) is the reduced Planck's constant, and is ~ 1.054E-34 Joules/second. Since this is a multiplication with a fixed result, the lower one goes (the less standard deviation and thus more precision), the higher the other one goes (the less we know about the other one) to maintain that value.
As a thought experiment, picture an atom whose electron you want to measure both the location and momentum of. You use a microscope for this. Say you want to know the exact position of the electron, you have to have a "high resolution". This is accomplished by using high-energy gamma rays (photons), since the max resolution you can get from a ray of wavelength λ is
sin θ = 1.22λ/D
where D is the size of the object. Therefore the smaller the wavelength, the better, since you want θ (the angle) to be smaller (the light will be less spread out).
So you shoot a high-energy photon to it and see where it bounces back to your microscope length. Problem is, now that it has bounced, you may get a very accurate picture of where the electron was, but you've changed its momentum by hitting it with a photon (Δx gets small, Δp gets big). This is because in QM the momentum of this gamma ray would be
p = h/λ
Where p is the momentum, and h is Planck's constant. So for a high energy photon, λ is smaller (shorter wavelength = more frequency), so p gets bigger, and you just disturbed the electron a LOT.
Similarly, if you wanted to go the opposite way and not affect its momentum so much, you'd use a lower energy means, but the image wouldn't be as "high resolution", so you'd have less knowledge of the position (Δx gets big, Δp gets small).
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Sun Aug 05, 2007 1:20 pm
I have slight reservations about the comment that the standard deviation "means how spread out the extreme values of a measurement are". The standard deviation is a measure of the general spread of the distribution and not the extrema. >_>;
The think with real problem with the HUP is that it is not readily intuitive. It is quite abstract because it relates the variance of the expectation values of a pair non-commuting quantum operators acting upon a state. neutral
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Sun Aug 05, 2007 8:31 pm
Unfortunately, the best description and explanation of HUP is the one that's most mathematical, and least intuitive. I mean, the mathematical formulation is the one that's *correct*; any other words or physical intuition that we attempt to attach to it can cloud our judgment or understanding of it.
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Wed Aug 08, 2007 2:48 am
It's weird. Momentum is energy in a direction, so doesn't something being in a direction imply position? And why doesn't this work like biliard balls? Given the streingth you hit the que ball and given where it stops, wouldn't you be able to tell via geometry and high-school physics where the ball that was hit was and what it's momentum was? I also imagined that if the billiard balls all traveled at the same speed and only varried in mass (ball mass I hope is a reasonable analogy to particle waveleingth) and it didn't seem to make anything undeterminable.
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Wed Aug 08, 2007 6:51 am
sodiumfree000 It's weird. Momentum is energy in a direction, so doesn't something being in a direction imply position? And why doesn't this work like biliard balls? Given the streingth you hit the que ball and given where it stops, wouldn't you be able to tell via geometry and high-school physics where the ball that was hit was and what it's momentum was? I also imagined that if the billiard balls all traveled at the same speed and only varried in mass (ball mass I hope is a reasonable analogy to particle waveleingth) and it didn't seem to make anything undeterminable. This is where the analogies with the macro world stop sweatdrop Billiard balls are well-defined things. You don't have any variation on x since they're pretty un-changing in position. Momentum does affect them, yes, but at the subatomic scale, everything is a particle and a wave. It's not only the momentum equation that applies to the electron, it's the position one too. You can freely observe the ball since a photon bouncing off it will have no impact on the ball's momentum, but this is not so for an electron. Simply observing the electron collapses its wavefunction - that is, makes it now stay in a specific state of position (the square of the values of the wave gives the probability distribution that the system will be in any given state). This is the intrinsic "uncertainty" in all measurements Heisenberg talks about - we modify an object when we measure it. At the macro scale, it's stupid to consider the effect of a photon bouncing off a billiard ball, but at the micro scale, "watching" an electron will disturb its states, since things aren't deterministic, but probabilistic. The same equations won't apply (sin θ = 1.22λ/D) to observing a ball since the energy supplied by a single (or, hell, a shower of) photons by p = h/λ is VERY small at the macro scale, and almost any electromagnetic wave can detect a billiard ball, there's no need to use an "ultra energy" photon to pinpoint its location. But this is all stretching the microscope analogy a bit too thin sweatdrop . Momentum would be m.v, the fact that v is "5 km/h at 30º" doesn't give you the position of it, just where (relative to itself) it will be at a given time, assuming constant velocity. Also the different particles travel at different speeds in vacuum - we all know light travels at c, but electrons travel at ~2.42 x 10^8 cm/s in the first atomic orbit, while it can move at 99% c when it's a beta ray. But honestly, the actual algebra that controls this thing isn't something I'm familiar with (for Christ's sake I haven't even seen group theory yet D:!), I only know the concepts behind it (you have two related variables that can't be measured at the same time with infinite precision), maybe Layra or Iggy could explain it a bit better? *non-subtle wink at them* heart Quantum mechanics: The dreams stuff is made of.
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Wed Aug 08, 2007 9:11 pm
*winks back* Your place or mine? . . . Or was it not that kind of wink? Anyway, the mathematics behind it requires that you take a couple leaps of faith before you get anywhere solid. Oh, sure, there's foundation all the way, but a lot of it is buried underneath what feels like loose soil. Believe me, everything here works but we don't have time for all of it. It boils down to a couple key ingredients: A system can be described by a wavefunction Y(x), which is just some complex function. You should think of this like a vector with a coordinate for every point x. To some extent, you can think of |Y(x)|^2dx as the probability that you'll find the particle at point x. An important point is that  Another important bit of notation is  where f(x)* is the complex conjugate of f(x). Note that = *
Observable quantities are associated with what are called operators that take the wavefunction and spit out a new function. For an observable (don't worry about the brackets, they just mean "average value" (kinda)), we associate the operator Q, and then we can calculate the value of = to be
 These operators are special, in that = . One thing to note is that if there exists a constant q such that Q(Y(x)) = qY(x) for all x, then = q.
The position operator is written as X, where applying X is just multiplying by x at every point. So X(Y(x)) = xY(x).
The momentum operator is written as P, where applying P is taking the derivative of the wavefunction and then multiplying by -ihbar, i being the square root of -1 and hbar being the reduced Planck's constant. The position operator looks kind of funny, but there's a reason for that, which I won't go into right now.
Now, if you could measure both position and momentum at the same time, you would be looking for something like
 Or, alternatively, you could look for
 And if this were a nice, kind world then we'd get the same things no matter which order we looked in.
But it's not.

In shorthand, we write (XP-PX)Y(x) = ihbarY(x). We can then write = ihbar, since (XP-PX)Y(x) is in fact a constant times Y(x).
"So what?" you ask. We should just pick one and have done with it. But it's not that easy.
Let's denote the standard deviation in position as , and the standard deviation in momentum as 
Given our notation, we get that the square of the standard deviation is
 which gives us
, We just put f = (X-)Y for shorthand. Similarly, we get that
 where g = (P-)Y for shorthand.
Thus
 which, by what is called the Schwarz inequality, must be greater than or equal to 
Now, for any complex number z, we get that |z|^2 = ((1/2i)(z-z^*))^2. Thus, letting z = , we get that

Now all we have to do is evaluate -, which, by some algebra, ends up as - = = ihbar. So we get that

Since standard deviations certainly can't be negative, we finally get that
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Thu Aug 09, 2007 1:20 am
Upon rereading the post I made above, I feel like someone ought to make an Intro to QM thread. Unfortunately, the very phrase "Intro to QM" screams of incompleteness and incomprehensibility. QM is nothing without the mathematics, and with the mathematics it is too much for a simple "intro"
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Thu Aug 09, 2007 7:24 am
I feel it's a tad unfortunate that position and momentum always get picked out for the uncertainty principle. So, the HUP is stated in terms of position and momentum, and is true for any canonically conjugate operator, but it can be generalized in a way that I think may help the concepts.
There are several things to note about Layra's mathematical exposition that are truly fundamental to quantum mechanics: 1) The state function Y(x) is inherently complex (in general). Everything in classical mechanics can be thought of in real space, but quantum mechanics doesn't have that luxury. 2) The equation for momentum is NOT p=mv in quantum mechanics (or even p=part(L)/part(qdot) ). It is p=-i*hbar*part/(part(x)), a result of the processes of quantum mechanics.
The crux of the issue is that position and momentum are no longer simply variables that you multiply in, like in classical mechanics, but are operators on the state function. In general, *operators do not commute with each other*. If you look through Layra's proof, you'll see that the noncommutativity of p and x are key points.
The general uncertainty principle (in semi-mathematical terms) is this: Two noncommutative observable operators have a lower limit on their simultaneous variations. Position and momenta are simply two such noncommutative observables, and the ones originally described by Heisenberg. This general principle points out two things: 1) The change from classical to quantum mechanics - from variables to operators, and real to complex functions - produces inherently different results (i.e. quantum mechanics is not trivial sweatdrop ) 2) There are observables which *can* both be measured with arbitrary certainty. The key is that they have to commute with one another as operators. For example, while (let's say in a free particle) position and momentum cannot be simultaneous eigenstates, obviously momentum and energy can, since (taking H=p^2/2m) H and p commute with each other.
I'll end my rambling here. I apologize for not knowing how to lay down TeX in Gaia.
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Thu Aug 09, 2007 10:18 am
For the TeX -> http://math.b3co.com and type there. You get the URL of an image-version of your LaTeX code. domokun *spends the next 3 days understanding the math in Layra's post* Yum knowledge. 3nodding
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Sat Aug 11, 2007 1:07 pm
Haha...I was about to say, "Layra, you should start an intro to QM thread like the intro to SR thread." I come back to write that, and there it is. That's hilarious.
On second thought, that's pretty creepy.
Rayquazza: thanks for the link. Do you have a link that teaches me how to put pics into posts lol?
|
 |
 |
|
|
|
|
|
|
|
|
|
|
Posted: Sat Aug 11, 2007 3:09 pm
Swordmaster Dragon Haha...I was about to say, "Layra, you should start an intro to QM thread like the intro to SR thread." I come back to write that, and there it is. That's hilarious. On second thought, that's pretty creepy. Rayquazza: thanks for the link. Do you have a link that teaches me how to put pics into posts lol? There's the IMG button up in the list of post-fun buttons. And as they say, "Great minds think alike." Which actually means that I don't think at all, I just steal other people's thoughts and pretend they are my own. In this case, your thoughts.
|
 |
 |
|
|
|
|
|
|
|
|
Posted: Wed Aug 15, 2007 9:16 am
So, that time I was working on a proof that the complex numbers are algebraically closed using winding numbers, and I knew how to prove it but lost it, causing me to lose half the points on the homework...that was you?
Grr. ::shakes fist::
(Man, Gaia really needs a ::shakes fist:: emoticon...I would use it in every post)
|
 |
 |
|
|
|
|
|
|
|
 |
|
|
|
|
|