« first day (1487 days earlier)      last day (3435 days later) » 

Sabre Tooth
Sabre Tooth
2:57 AM
stupid question - but would a question about a practical application of a Faraday cage in a classroom be on topic?
 
Brandon Enright
Brandon Enright
3:32 AM
@SabreTooth As long as the question is about a concept and isn't just looking for a list of experiment ideas or similar then sure, it would be on topic.
 
user54412
I'm calling this early - there is some great physics lurking behind physics.stackexchange.com/questions/149580/…
 
user54412
I'm not the one to do it, but there's definitely plenty of thermodynamics and kinetics effects mediated by fluid dynamics and other microphysics to produce that effect (which I've seen for myself)
2
 
Sabre Tooth
Sabre Tooth
5:02 AM
@BrandonEnright i'll think about how to write for when I can ask a question
ah wait a second, it would have been experimental applied physics - so that is off topic... what a shame
 
hwlau
hwlau
@SabreTooth If you are only asking about physical principle, then it would be on topic
 
Sabre Tooth
Sabre Tooth
5:19 AM
I am thinking of the question about 'mesh faraday cages' and asking if there is a relation between mesh size and how much of the EM radiation the cage blocks
 
dmckee
dmckee
@SabreTooth Experimental questions are on-topic.
@SabreTooth There is a perfectly fine undergraduate theory question there, or a deeper one where a bunch of undergraduate assumption are relaxed; and also a practical experimental question as well.
 
Sabre Tooth
Sabre Tooth
@dmckee so, a good question, potentially?
 
dmckee
dmckee
The only thing is that you have to ask a question, not just throw it out there wide-open and unspecified and hope that someone can guess the what question you really mean.
 
Sabre Tooth
Sabre Tooth
oh the question will be a single one
Is there a relation between mesh size and the amount of a phone signal is blocked
@dmckee would that be focussed enough?With a question title of "Can a mesh Faraday Cage completely block cell/mobile phone reception?"
 
dmckee
dmckee
@SabreTooth That is exactly the kind of question that a layman would view as well specified, but drives the pros bats. Define "completely"? Is 30 db good enough? 50? Also helpful if you define the band involved and the mesh size you have in mind.
The best mesh Faraday cage I ever saw lived for a while in the high-bay at KSU's physics department (I never learned why, we'd never used it). The mesh had a spacing of around 3 mm.
 
Sabre Tooth
Sabre Tooth
5:32 AM
hmmmmm
that's drifiting away from what I am wanting to ask
okay @dmckee what I mean by 'completely' is 0dB - i.e. completely blocked
@dmckee or maybe better still "Is there a relationship between Faraday cage mesh size and the amount of UHF signal blocked?"
@dmckee nevermind, appears to be many similar questions
and besides, my only question here still does not have an answer, despite an earlier bounty... I know experimental physics is on topic here, but it does not seem well received (and I am most definitely not a theoretical physicist)
sorry to have bothered you
 
 
2 hours later…
David Z
David Z
7:20 AM
@SabreTooth yeah, that's because there aren't a lot of people with experimental expertise here to answer those questions. We're always thinking about how we can improve participation in the site, so if you have any ideas about how to get more experimentalists to participate, please do share!
well, not just getting them to participate, but getting them to stick around. We need to somehow get a critical mass of experimental physicists. "somehow" is the tricky part :-P
 
Brandon Enright
Brandon Enright
@DavidZ We're fortunate to have Anna V for an experimentalist's view but more would be nice.
 
Sabre Tooth
Sabre Tooth
I am not just an experimental physicist,I also dabble in instrumentation (design and development, and a lot of time re-engineering)
 
Brandon Enright
Brandon Enright
I do think you can ask a good on-topic question about cell phones and Faraday cages.
 
Sabre Tooth
Sabre Tooth
it seems it would be a duplicate
I am interested in mesh size vs. signal reception
 
Brandon Enright
Brandon Enright
@SabreTooth I bet it's a step-function. Mesh size greater than the wavelength would probably have very little attenuation and mesh size smaller would have very high.
But that's just my guess.
 
Sabre Tooth
Sabre Tooth
7:31 AM
it seems such a question would be a dupe of physics.stackexchange.com/questions/89584/…
okay, I'll try somehing
 
Brandon Enright
Brandon Enright
None of those really focus on the hole size and what happens when the hole size nears the wavelength size.
Your question would be related to the ones you listed but I think you can ask it in a way that it's both interesting, conceptual, and not a duplicate.
 
Sabre Tooth
Sabre Tooth
7:58 AM
@BrandonEnright done, I hope .... physics.stackexchange.com/questions/149607/…
writing that was kind of ffun.... but now I need cookies
oh, and please feel free to edit anything that is mispelt, not clear etc
hello @ManishEarth
 
ManishEarth
ManishEarth
o/
 
Sabre Tooth
Sabre Tooth
with the kind help of @DavidZ and @BrandonEnright I was able to compose a symphony of a question
 
 
4 hours later…
David Z
David Z
12:03 PM
1
Q: Some Thai followers say our mind is faster than light?

PantipThe sun's light takes 8 minutes to reach the Earth but our mind can think of the sun or even of distant stars instantly. That is to say our mind is faster than the speed of light. They always claim that Sciene is the subset of Buddhism. What do you think about these notions. true or false?

science lay-buddhism
hehe
 
 
3 hours later…
Jim
Jim
3:03 PM
Re David's post: I did once read somewhere years ago that some scientists think the brain utilizes quantum entanglement to send signals between the left and right halves. I'm not sure how credible those beliefs are, but maybe nature found a way to send information via entanglement where we could not.
But who am I kidding? That's utter nonsense
 
 
2 hours later…
user54412
4:36 PM
After all the hassle I just went through to get 30 year old Phys. Rev. articles, I'm glad to be in astrophysics, where all research to speak of is freely available after a year at most.
 
Qmechanic
Qmechanic
4:51 PM
@Phonon and other commentators: Concerning use of lmgtfy e.g. here, see also this meta post. Google search links are fine and helpful but lmgtfy links are discouraged.
 
Danu
Danu
5:08 PM
@Qmechanic I agree that it's not good to use those; they are kinda mean.
 
 
2 hours later…
Danu
Danu
7:03 PM
I've got a random question
 
dmckee
dmckee
We've got random answers.
 
Danu
Danu
how can I find a solution for $2x^3+3x^2+x-3N=0$ for large $N$?
Can I set up an expansion somehow?
In $N^{-1}$ I presume
but I have no idea how to get it done, really
 
dmckee
dmckee
There exist closed form solutions for 3rd degree polynomials, but they are a hassle.
Try googling it?
 
Danu
Danu
Yeah, and they don't help. I've been working on this for over an hour :(
 
ACuriousMind
ACuriousMind
@Danu Uhhh...all third order polynomials are solvable - why don't you just plug it into Cardano's formula?
 
dmckee
dmckee
7:05 PM
I'm told that there are also closed form solutions for 4th degree, but not above that.
 
Danu
Danu
As it turns out, Mathematica cannot even find out whether any of its roots are real
@ACuriousMind $N$ is a variable
 
dmckee
dmckee
The 4th degree one is horrific. The algorithm spans several pages.
 
Physics Meta
Physics Meta
0
Q: Are questions asking for confirmation of a method off topic?

FraserOfSmegI have asked a question relating to how I have approached a problem. My reason for asking the question was my answers seem to be unrealistic. The question is here: How to calculate force on a string pinned to a spinning object It was put on hold reasonably soon after I asked the question. I jus...

discussion
 
dmckee
dmckee
@ACuriousMind Well, having the name will help with googling it.
 
Danu
Danu
@ACuriousMind I cannot determine which of its solutions are real (or whether they all are, one of them is, etc)
@dmckee I already knew it, and I already did that ;) Didn't help
 
ACuriousMind
ACuriousMind
7:07 PM
@Danu It has to have a real root, since the complex numbers have only degree two over the reals
 
Danu
Danu
@ACuriousMind Yeah, but which one ;)
In fact, I'm fairly certain I found a real solution
but the series expansion of mathematica 'features' $(-1)^{2/3}$
:(
(as does it for both other roots)
 
ACuriousMind
ACuriousMind
Hmmmm
Very helpful :D
 
Danu
Danu
Very strange, mostly :\
^lol
dat realization ;D
 
user54412
@Danu Numerical recipes section 5.6
 
Danu
Danu
@ChrisWhite hulp pliz me no understend how do solve?
 
user54412
7:11 PM
$x^3 + a_2 x^2 + a_1 x + a_0 = 0$, $a_i$ real, searching for real root, right?
 
Danu
Danu
pretty much
as $a_0\to -\infty$
that's important
I'm looking for an asymptotic expansion
 
user54412
define $q = (a_2^2 - 3 a_1)/9$, $r = (2 a_2^3 - 9 a_1 a_2 + 27 a_0)/54$
 
user54412
if $r^2 - q^3 < 0$, then define $\theta = \cos^{-1}(r/q^{3/2})$, and the real root is $-2 \sqrt{q} \cos(\theta/3) - a_2/3$
 
Danu
Danu
it's not
$r\to \infty$
oh wait
no sorry
$a_0\to -\infty$
 
user54412
otherwise define $a = -\sign(r) (\abs{r} + \sqrt{r^2-q^3})^{1/3}$, $b = q/a$ ($b = 0$ if $a = 0$), and then the real root is $a + b - a_2/3$
 
user54412
7:15 PM
all real operations, no diving into $\mathbb{C}$ just to end up with a real result in the end
 
Danu
Danu
mhm
that seems helpful
 
user54412
I only know this since the code I'm working on has a routine that relies on explicit cubic (and quartic!) root finding
 
Danu
Danu
let me implement :)
 
user54412
hopefully I didn't mistype -- it should be Numerical Recipes, 3rd ed., equations 5.6.9-5.6.18
 
Danu
Danu
Damn... this also isn't helping much
 
ACuriousMind
ACuriousMind
7:20 PM
@Danu: What kind of assignment does this belong to?
 
Danu
Danu
mathematical quantum mechanics
 
Qmechanic
Qmechanic
@Danu : Assuming that N>>1, solve it as a recursive fixed-point equation x:=((3N-3x^2-x)/2)^(1/3). Start with seed x=0 on RHS. First iteration yields x=((3N/2)^(1/3), next approximation yields x=((3N/2)^(1/3) - 1/2. A couple of iterations later: x=((3N/2)^(1/3) - 1/2 + (1/12)(3N/2)^(-1/3), and so forth.
 
user54412
there's a way to do the calculation for quartics as well, similarly keeping track of branch cuts so as to always work with real numbers, but almost no source on the internet will tell you about it
 
Danu
Danu
@ChrisWhite thanks a lot for your help, but I don't think this is really getting me anywhere hha
The series expansion is quite terrible
@Qmechanic I'm not sure how to implement this (I am terrible at numerical methods)
@Qmechanic The problem is that $N$ is a variable, I'm looking for an asymptotic expansion
in particular, I cannot fix $N$ to be some number
 
Brandon Enright
Brandon Enright
Mmmm Numerical Recipes, one of the best books ever.
 
user54412
7:33 PM
@Danu I think Qmechanic's method works if you just take N to be much greater than $x$ in his equation
 
user54412
At least, it gets the same result I get for assuming N large in my method
 
Danu
Danu
@ChrisWhite I need to get the $N$-dependence (i.e. some function of $N$, which I can then expand)
Also, I'm a total n00b at numerical methods so if I'm missing something obvious let me know, please
 
user54412
well, (3N)^{1/3} looks like N-dependence to me ;)
 
user54412
or do you want even more terms
 
Danu
Danu
@ChrisWhite I'd like to have something more than throwing away everything but $x^3$
I suspect there is a constant term
 
user54412
7:35 PM
@BrandonEnright Agreed. One of those books everyone should have next to them at all times
 
Danu
Danu
@ChrisWhite Could you give the author etc? I need to have some numerical methods reference! :)
 
user54412
Numerical Recipes
Numerical Recipes is the generic title of a series of books on algorithms and numerical analysis by William H. Press, Saul A. Teukolsky, William T. Vetterling and Brian P. Flannery. In various editions, the books have been in print since 1986. The most recent edition was published in 2007. == Overview == The Numerical Recipes books cover a range of topics that include both classical numerical analysis (interpolation, integration, linear algebra, differential equations, and so on), signal processing (Fourier methods, filtering), statistical treatment of data, and a few topics in machine learning...
 
user54412
it even has a wiki page (though I guess lots of books do these days)
 
user54412
It's not the final word on anything, but it's a solid introduction to just about every numerical technique out there.
 
Danu
Danu
@ChrisWhite thanks a lot :)
@Qmechanic I think I understand the idea, but I'm not sure how to program this
I also tried to work the 2nd iteration out by hand, and dont see how to simplify to 3/2
@Qmechanic One should use $x_1=(3/2\ N)^{1/3}$, and then apply the fixed point formula to find $x_2$, right? I don't see how to simplify
 
Qmechanic
Qmechanic
7:54 PM
@Danu : Taylor expand treating 1/N as small.
 
Danu
Danu
@Qmechanic Alright, that sounds like what I was trying to do earlier (in a different way). I'll give it a shot
 
Danu
Danu
8:06 PM
Hm... I just ended up doing a recursive taylor series expansion
which, funnily enough, doesn't give me this constant term
I find $x\approx (3N/2)^{1/3}(1-(3/2)^{2/3}(3N^{1/3})^{-1}-(3/2)^{1/3}(9N^{2/3})^{-1}+\dots)$
@Qmechanic isn't that strange?
 
Danu
Danu
8:24 PM
@Qmechanic I really don't seem to be able to get the $-1/2$ out of my series...
 
dmckee
dmckee
Note that Numerical Recipes has had some persistent errata and has sometimes been slow to pick up on new algorithms. But it's a great book for all that.
 
Qmechanic
Qmechanic
Mathematica: In[1]:= f[x_]:= (y^3-(3x^2+x)/2)^(1/3)
In[2]:= Series[f[f[f[f[f[f[0]]]]]],{y,Infinity,6}] (Here y:=((3N/2)^(1/3).)
 
Danu
Danu
@Qmechanic Interesting, thanks. Is it immediately clear to you what's wrong with the following:
$x\approx (3N/2)^{1/3}\left(1-\frac{x}{9N}-\frac{x^2}{3N}\right)$
and then using only the first term of the expansion of the $x$'s in the ratio's, since the second one will be at least $O(N^r)$ with $r>1$
 
Danu
Danu
8:52 PM
I still don't understand why the constant doesn't show up that way.
 
Danu
Danu
9:09 PM
@Qmechanic In fact, this gets more interesting when one considers that the 'official solution' does not seem to count on the constant being there. I really wonder what's going on here
 
Qmechanic
Qmechanic
@Danu : Please double-check that the 3rd order equation is written correctly.
 
Danu
Danu
@Qmechanic The initial form is $2K(\frac{K^2}{3}+\frac{K}{2}+\frac{1}{6})=N$
Multiplying by three gives the equation I gave, right?
$2K^3+3K^2+K=3N$
 
Qmechanic
Qmechanic
@Danu : Yes.
 
JamalS
JamalS
@Danu: What did you try in Mathematica? I was able to get an expansion for large $N$.
 
Danu
Danu
@JamalS Nice. With the help of Qmechanic (who basically gave me everything) I was able to as well
I'm just intrigued by the presence of this 1/2
which is not there when I try this 'recursive Taylor expansion' type thing
and also doesnt seem to be assumed in the solution
 
JamalS
JamalS
9:27 PM
@Danu: Out of curiosity, why did you need to solve that equation?
Which problem gave rise to it?
 
Phonon
Phonon
10:00 PM
@Qmechanic sure, thanks for letting me know, didn't know!
this opencourseware movement is really taking on
MIT making more and more lectures available online for free, same with Stanford. Now Oxford's starting as well.
so so nice
 
Danu
Danu
10:20 PM
@JamalS Sorry for the late reply, had to make dinner (so late... no time these days)
It arose in a problem that I'm doing for my class in math. QM
The (sub)question is literally telling me to invert the qubic equation to $o(1)$
Based on the subsequent questions, I can see what the answer is supposed to be
but the constant is intriguing
 
Physics Meta
Physics Meta
11:11 PM
0
Q: reviewing first post

tomIs there any guidance on reviewing the first post of a new user? I have looked in meta and could not find a similar question. All I can think of at the moment is to compare with guidance on the quesiton help here if it is a question and compare with guidance on answering if it is an answer and...

discussion review new-users
 

« first day (1487 days earlier)      last day (3435 days later) »