« first day (2508 days earlier)      last day (2720 days later) » 
01:00 - 13:0013:00 - 00:00

01:40
@ACuriousMind Suppose I have a matter field $\psi$ in an EM field $A$, and the Lagrangian is $$\mathcal L(\psi, A)=|D_\mu \psi|^2-\frac{m^2}{2}|\psi|^2+\frac{1}{p}|\psi|^p.$$ Is the $p$-nonlinearity anything physical or just a nonlinearity for nonlinearity's sake?
I think chemistry will start to become electron shell physics when atom by atom synthesis of molecules become possible
Imagine being able to place atoms precisely in space and then they form a molecule. No more complicated organic synthesis procedure needed
I am starting to wonder, whether mathematics can demonstrate the existence of such direct path of synthesis for any complex molecules...
This is because viewing chemical synthesis in a mathematical perspective, it forms a configuration space with reaction conditions as coordinates where paths can be drawn
02:37
@Secret There will be a lengthy time when you 'can' build molecules to order but organic chemists and chemical engineers continue to do what they have always done.
Because a mole is such a big number.
On the other hand, tailored biological sourcing could sweep the industry very fast if it turns out to be easy to grown and maintain large cultures
 
1 hour later…
Anonymous
03:50
@0ßelö7 "I say it all the time"<----that is not a reason for it to be correct in the context you used it. A "grammatical rule" could be any rule which conforms to the rules of grammar. Even "Sum of two natural numbers is always greater than the difference of the same two same natural numbers. " is a grammatical rule as it conforms to the rules of grammar. On the other hand "grammar rule" is a "rule of grammar", and is much more widely used.
Anonymous
1
Q: Which is correct: "grammar rules" or "grammatical rules"?

Blue Which is one is correct English: "grammar rules" or "grammatical rules" ? In my opinion, "grammatical" means "conforming with the rules of grammar". Thus, "grammatical rules" should literally mean "rules which conform to the rules of grammar". That is surely not the definition of "grammar...

Anonymous
1
A: Which is correct: "grammar rules" or "grammatical rules"?

RingoYes, I think you are correct! Grammar rules means rules about grammar. Grammatical rules are rules that are well-written and grammatical. I think this means you can have grammatical grammar rules. In fact, grammar rules should always be grammatical!

eye roll
Anonymous
Anyhow, you need to get rid of the notion that your customs are the laws of nature.
nope
that's a good way of putting it
"law of nature" -- nice!
Anonymous
03:55
I don't see any counter-argument there, so I suppose we're done.
04:12
grammar rules is too damn cringy though
@0ßelö7 Well, that works only if $\Omega$ is convex.
But yeah there's some technicalities about smoothing as you approach limit; it needs to be done carefully
@Kaumudi.H Ted has a series of lectures on multivariable calculus, part of a course out of his book (from which I learnt multivariable calculus). Might check it out at some point.
@BalarkaSen close enough to the boundary we can assume it's convex
I clearly only care about the construction for large $n$...so just cut off the sequence
Ok, yes, true.
@BalarkaSen Hmm. I don't need it to be $C^1$ up to the boundary, just $C^1$ on $(0,1)$, say
user228700
@BalarkaSen Ohh, where (if) are they available to watch?
And the curve needn't be embedded, self-intersections are allowed
04:19
@Kaumudi Look in his profile description; they are all uploaded to youtube
user228700
Ah, oops. Thanks!
@BalarkaSen Basically I want to estimate $\lim_n f(x_n)/g(x_n)\to 0/0$ and L'Hopital works very well if I can instead do the limit along a $C^1$ curve
Hmm. I might need $C^1$ up to the boundary for L'Hopital...
Amazing
You don't need the functions to be differentiable at that point
I have seen something like a C^0 L'Hoptial before in a proof by Thurston
@BalarkaSen So I think what I described is a legitimate inductive construction. I don't think smoothness could go wrong there.
I'll believe you!
user228700
04:31
Sigh, it's 10 o'clock on a lovely Saturday morning and this, this is what I'm doing:
user228700
04:46
@Kaumudi.H My sympathy is limited because at your age I was absolutely fascinated by computers and would happily have spend every waking moment playing with them.
^sad
@0ßelö7 no, I was very happy as a teenage computer nerd. It was teenage computer nerds like me (but with more ambition) who built todays computer industry.
And to be honest in the UK in the late 70s there was frack all else to do :-)
@JohnRennie golden age of global calculus
you could have done that
user228700
I have a quick question! Do two waves need to be coherent to interfere? Not to interfere completely constructively/destructively, merely to interfere.
Anonymous
@Kaumudi.H Define coherence and you'll get the answer
user228700
04:59
Huh, well...
Anonymous
Mathematically, the answer is "yes"
Anonymous
But practically speaking there's something called "degree of coherence"
user228700
Clearly, this is not a situation in which having little information about the definition of coherence is helpful :-/
Anonymous
@Kaumudi.H Did you get your answer?
user228700
Only sort of.
05:02
@Kaumudi.H any waves will intefere, coherent or not, but to get a stable interference pattern you need coherent waves.
Interference is just vector addition of the waves.
user228700
^ This was my initial guess as well :-/
user228700
I was confused for a moment only because my textbook has defined it differently, to suit the context.
Anonymous
For sustained interference you need them to be coherent
Anonymous
But, yeah
@Blue for a stable interference pattern you need them to be coherent. I'm not sure what sustained interference means in this context.
user228700
Heh, I have never liked this topic :-/
Anonymous
It's same as stable interference
Anonymous
:P
user228700
@JohnRennie :-) Well, not me!
Anonymous
@Kaumudi.H This topic forms the basics of QM
Anonymous
05:07
It's actually very interesting
@Kaumudi.H we can't all be perfect :-)
user228700
Well, I've never been able to form an intuitive understanding of it, hence the dislike.
user228700
@JohnRennie :-)
user228700
Anonymous
@Kaumudi.H Don't worry. Even the elite physicists of the 19th and 20th century were boggled by this. :P :)
user228700
05:08
Itza today!!
Anonymous
Wow, cool
user228700
@Blue Yeah :-)
@Kaumudi.H Have fun! It looks like it will be a great time.
user228700
@JohnRennie Well, I hope! :-) Thanks!
user228700
*INCLUSIVE OF DINNER (Huzzah!!)
Anonymous
05:14
@Kaumudi.H It's not free. You paid for it (indirectly). :'Dd
Anonymous
Do enjoy :)
user228700
No, it's not, haha :-) STILL! DINNER! At a TED event! :-P
user228700
Thank you :-) What are your plans for the weekend?
Anonymous
@Kaumudi.H I'm going out for lunch with family and relatives (uncle, aunt and cousin) today. :D
Anonymous
Then I'll go for Durga puja shopping
user228700
05:18
Oh, nice! :-)
user228700
@Blue Ah, when is it, again?
Anonymous
@Kaumudi.H Starts from 23rd I think
Anonymous
I'm not interested in the "puja" part though :P
user228700
Cool!
Anonymous
Only the food
Anonymous
05:18
lol
user228700
Hehe :-)
user228700
Oh, BTW, this is the theme of the event:
user228700
user228700
It's a bit vague, but we'll see!
user228700
05:41
Can anybody please help me with the following equation:
user228700
$l_o={{\lambda}_o}^2/\delta\lambda$
user228700
Where $l_o$ is the coherence length, $\lambda_o$ is the wavelength of the light in question and $\delta \lambda$ is the wavelength spread.
user228700
How does this work?
Suppose you have light made up from two wavelengths $\lambda$ and $\lambda+\delta\lambda$.
user228700
OK...
05:46
And they start in phase.
user228700
OK.
Because their wavelengths are different they will get out of phase after a certain number of wavelengths.
user228700
Hmm.
Suppose we kind of wave our hands around and argue that they will get out of phase when the difference is one wavelength. The number of cycles this takes is $\lambda/\delta\lambda$.
user228700
Right.
05:49
e.g. if they differ in wavelength by 1% then it takes 100 cycles for them to differ by one wavelength.
user228700
Ah, right.
The distance is takes is the number of cycles, $\lambda/\delta\lambda$ times the wavelngth $\lambda$, hence the formula you cite.
user228700
AHHH.
user228700
I s'pose I must let go of any hopes to understand what that hand waving involved, exactly :-P Though I'm satisfied with this much!
Of course in reality we have a continuous spread of wavelengths and $\delta\lambda$ would be something like the half width of the wavelength distribution.
So it's all a bit hand waving, and the coherence length is somewhat vaguely defined.
user228700
05:51
@JohnRennie Huh?
user228700
I wasn't able to parse that properly. What dyou mean, exactly?
In real life light isn't a single wavelength $\lambda$. If you spectrum analyse it you get a spread of wavelengths centred on $lambda$.
user228700
Ah, right, right, OK.
So what I'm saying is the width of this wavelength distribution would be what $\delta\lambda$ means.
user228700
Yes, it says:
user228700
05:55
> "The emission will not be strictly monochromatic and will have a wavelength spread (spectral broadening) due to a variety of reasons such as the inevitable width of the energy levels and the Doppler broadening due to random motion of the emitting atoms and molecules"
user228700
$=\delta \lambda$
user228700
Alright, thanks so much! :-)
user228700
@JohnR: Are you working ATM?
Kind of, but I have attention to spare for physics if you want to ask anything.
user228700
Yes, I was wondering if we could complete that discussion on DSHO but I'm not too sure what I'm looking to understand, really, because, see, this is what my textbook is like:
06:07
Yes, do you want to do it in the deep end or here?
user228700
user228700
AH, the amount of time that took!
user228700
I dunno what to do, really, whether to learn this by heart without understanding a thing or to spend hours upon hours trying to.
I would suggest we take a step back and let me explain how we approach the DSHO. I think the time taken will be worth it in the long run.
user228700
@JohnRennie OK.
06:11
Ok, you remember how I approached the undamped SHO? I guessed the solutionwas $e^{i\omega t}$, fed this into the differential eqaution and showed that (a) it was a solution and (b) we got the value of $\omega$.
user228700
Yes, I remember.
OK, we adopt exactly the same strategy for the DSHO.
user228700
OK...
In the limit of no damping the solution has to be $e^{i\omega t}$, so my guess is that with damping the solution will be $e^{i\omega t}$ times some decay function e.g. an exponential decay $e^{-kt}$. Does this seem sensible?
user228700
Yes, absolutely.
06:16
So the solution is going to be $$x(t) = e^{-kt} e^{i\omega t}$$
user228700
Hmm, yes...
And a bit of rewriting gives $$x(t) = e^{(-k + i\omega)t}$$
user228700
Yes...
Or: $$ x(t) = e^{at}$$ where $a$ is a complex number
user228700
Ah, right.
06:19
So what we do is feed this into the equation for the DSHO: $$\frac{d^2x}{dt^2} + \frac{k}{m}\frac{dx}{dt} + b^2x = 0 $$ (I think that's all correct though it's worth checking)
Are you happy with this so far?
user228700
Sort of...
Oops, just spotted a sign error
user228700
Where?
@Kaumudi.H you need to be 100% on this, because it's our starting point
@Kaumudi.H I've just corrected it.
Looking at your book it gives the equation as: $$\frac{d^2x}{dt^2} + \frac{\gamma}{m}\frac{dx}{dt} + \frac{C}{m}x = 0 $$
Equation 1.2
user228700
@JohnRennie I'll tell you where this completely loses me, John, it's in the physical significance of all these friggin' constants. I remember what you said about the constants serving a purpose while solving it etc. but still.
06:25
Do you mean how we got that differential equation in the first place?
user228700
No, no, I understand that.
Do you mean how the $\gamma$ and $C$ are going to relate to things like the frequency?
user228700
YES, yeah.
That's what we'll discover when we feed our guessed solution back into the equation.
Shall I do that now?
user228700
OK, let's see, then!
user228700
06:27
I'll try on the side.
It's probably best if you do this on paper along with me.
user228700
34 secs ago, by Kaumudi. H
I'll try on the side.
user228700
:-)
We start with the guess $$x(t) = e^{at}$$
So: $$\frac{dx}{dt} = a e^{at}$$
user228700
Hang on, hang on.
06:29
OK?
user228700
OK, so I jumped a few steps ahead (:-P) and so, $e^{at}$ can't be zero?
user228700
No, no, it can't, what a stupid question. OK.
So where have you got to?
user228700
So the other expression is zero
user228700
$(a^2+\gamma a/m + c/m)=0$?
06:32
Hang on, let me write it out ...
Yes, I get the same as you. So we have a quadratic equation for $a$ and all we have to do is write down the solution.
user228700
I get some contorted expression for the two solutions for $a$
I get: $$ 2a = -\frac{\gamma}{m} \pm \sqrt{\frac{\gamma^2}{m^2} - \frac{4C}{m}} $$
Do we agree?
user228700
Nope. Hang on...
user228700
Yeah, yeah we agree. I'm dumb. Carry on!
OK, we'll do a quick sanity check. In the undamped limit we have $\gamma \to 0$. Yes?
user228700
06:38
Tends to?
user228700
Why?
If $\gamma = 0$ there's no damping ...
user228700
Right, and no damping = undamped, no? :-P
user228700
Right, so my question is why it only tends to zero, why it isn't zero.
06:39
You're over thinking my statement.
Anonymous
If it were 0 it wouldn't be DSHO :P
I just meant as we gradually reduce the damping to zero wer have $\gamma$ tending to zero.
user228700
@JohnRennie :-/ I'm sorry. OK.
user228700
Right.
Anonymous
If the $dx/dt$ term vanishes, then it would be your ordinary SHO
06:40
OK. So put $\gamma = 0$ into our solution and what do we get?
user228700
@Blue Yep, yeah, I get that.
Anonymous
You can think of it as the damping coefficient. More is the $\gamma$ more is the damping.
@Blue please don't distract K
user228700
@JohnRennie C is still significant?
user228700
@JohnRennie Lol.
06:43
Just take our solution: $$ 2a = -\frac{\gamma}{m} \pm \sqrt{\frac{\gamma^2}{m^2} - \frac{4C}{m}} $$ and put $\gamma = 0$
user228700
Yes, right, since there is a -4C/m term inside the square root if we do that, I was wondering if...ah, see, I forgot something important in the middle, never mind.
Focus young lady, this is just algebra
user228700
Yeah, yeah, sorry :-( So $k=0$ and $\omega t= \sqrt{C/m}$ ?
user228700
Oh, ah, that $\pm$ Didn't account for that.
There's a rogue $t$ in there, but yes I get $a = i\sqrt{C/m}$, and our solution is $e^{at}$ so the solution is $e^{i \sqrt{C/m} t}$.
user228700
06:48
Right, sorry about that $t$
user228700
Give me 2 mins?
But you've got the point I think. What we've shown is that $\sqrt{C/m}$ is the frequency of the undamped oscillator.
user228700
We have. WE HAVE!
user228700
Where did $k$ go though?
Your book uses $\gamma$ not $k$, so I did the same.
user228700
06:50
::Facepalm::
user228700
No, no.
user228700
::Head-desk::
:-)
This is the thing. It's actually really straightforward when someone explains it to you.
user228700
You, when you explain it. My teacher? Yeah, he also tried.
Once we got our equation for $a$ lots of other things are obvious. For example the sqrt is imaginary if $$ \frac{\gamma^2}{m^2} < \frac{4C}{m} $$
and it's real if $$ \frac{\gamma^2}{m^2} \ge \frac{4C}{m} $$
user228700
06:53
Yep, yeah.
So we immediately get the condition for critical damping
Which is just $$ \frac{\gamma^2}{m^2} = \frac{4C}{m} $$
user228700
Wait, OK, so this is the thing: none of my devices have any charge left, I'm panicking for other reasons and I'm also standing here charging my devices, panicking, which is why I was so distracted and dumb (and I'm sorry) before.
user228700
So now I have to calm down and this will take me a solid 10 minutes of work.
user228700
And then after that, I gotta write a letter or something, to go out tonight, to the event. To the warden, remember I told you about me having to write a letter to go out.
I'd take a few minute out to write the letter now. Then have a coffee, then come back to this.
It's a very common reaction to a page of maths that looks impenetrable. We all react that way.
user228700
06:57
Without the option of coffee. I don't have coffee. Nobody has coffee.
user228700
@JohnRennie No, no, it's not the Math :-P
user228700
Anyway, right, I will be back, all calm, in 10 minutes.
WHAT? Give me your address and I'll go onto Amazon and buy you some coffee :-)
No coffee? YOU COULD DIE!!!! :-)
user228700
:-)
user228700
NO COFFEE! OK, the agenda was the stop panicking :-P
user228700
06:59
10 minutes. 10.
user228700
07:12
@JohnR: I'm back :-)
user228700
@JohnRennie Yes.
So where did we get to?
user228700
I suspect that you were going to talk about the three cases that arise.
It's probably not worth going any further with this because I suspect it's obvious to you now ...
The point I wanted to make is that now we have all the info we need to deduce the various things about how the DSHO behaves.
user228700
07:15
Obvious would not have been my word of choice :-) Though you are correct that I will be able to move on now.
user228700
@JohnRennie Yes, indeed.
OK. Ping if you want to ask anything further. I'll go back to kicking my servers :-)
user228700
I have been saved from the hells of absolute incongruity and utter confusion! YES! Thank you :-)
Anonymous
@JohnRennie Don't break them :P
Anonymous
@Kaumudi.H farrago XD
07:18
To be fair, if you look back at your book you'll see that it does exactly the same as I've just done, only with less explanation.
user228700
@JohnRennie Yes, I also noticed that the key difference is that you didn't assume the value of $\omega$ in terms of $C$ before you started, while my textbook did.
@Blue I'm working remotely, otherwise I'd be strongly tempted to start throwing the servers out of a high window :-)
user228700
This was incredibly confusing to me at the time.
user228700
@Blue :-P
@Kaumudi.H yes, the book jumped ahead. If I was writing the book i wouldn't have done that! :-)
user228700
07:20
:-) You're only 56. Quit kicking servers and sit down to write a textbook for us dummies!
user228700
I'll be going now. Must have lunch, write that letter, figure out a way to reach the event, get dressed, and, well, leave!
user228700
(Lunch, if you're wondering, is the same as breakfast today: bread with nutella)
user228700
Thanks so much :-)
user228700
I'll see you later. ::Waves::
08:16
What's the relaxation time?
@Blue ^^
Anonymous
@WrichikBasu Time between two successive collisions
Anonymous
Of electrons
Anonymous
Approx time
@Blue so, the more the relaxation time, more is conduction, right?
1, 3 and 4 are correct.
If the mass is more, the collisions would tend to increase, resulting in less conduction. So, 2 is incorrect.
@Blue ^^
 
1 hour later…
09:23
@Kaumudi.H If you're interested, David Morin's classical mechanics book has a very good chapter on oscillations. He does all those differential equations and goes through each case thoroughly
09:41
@JohnRennie can you help me in the above sum ^^
Please see the last pic posted by me here.
I think it's just 3
The question asks what is proportional to the applied field.
The mobility is a fundamental property of the material so it wouldn't be proportional to the voltage.
Likewise the effective mass of the charge carrfier.
@JohnRennie I see.
And the relaxation time is the time the electron takes to come to a halt when the potential is removed.
The relaxation time is certainly important in conduction, but I don't think it is proportional to the applied field.
@JohnRennie Oh, then that won't affect.
The point is that the question isn't asking which of the four affect conduction. It's asking which of the four are proportional to the applied field.
09:45
@JohnRennie yes.
I think only drift speed will be the answer. As you've said.
What about this one? ^^
I think 1 and 2 are correct. But the temp is high. That makes me think a bit.
3 may also be correct, as the energy gap is proportional to $kT$ where $k$ is Boltzmann's constant.
@JohnRennie ^^
I have no idea what the state of silicon is at 5000K. What is the boiling point of silicon?
3265°C
Less than 5000K.
Ok so the silicon atoms are probably present as single atoms so they won't be $sp^3$ hybridised so (2) is wrong.
And since the silicon is a gas not a liquid or solid there are no energy bands. So (3) is wrong.
How does kT at 5000K compare to the ionisation energy of silicon?
@JohnRennie I don't know. I didn't even know that Si would be liquid at the Sun.
Well according to Wikipedia the first ionisation energy is 786.5 kJ/mol. Since that's per mole not per atom we need to comapre it to RT, which is 8.31*5000 = 41.55kJ.
So the temperature isn't high enough to ionise the silicon.
09:57
@JohnRennie ok.
So we have a gas of neutral silicon atoms, so (4) is wrong.
That just leaves (1)
@JohnRennie understood.
In a semi conductor, does effective mass of majority career affect conduction?
shouldn't the energy difference between 3p energy levels always be rather small :o
@WrichikBasu Yes. The force on a charge carrier is $F = qE$ where $q$ is the charge (normally $e$) and $E$ is the electric field. So the acceleration of the charge carrier is $a = F/m = qE/m$, where $m$ is the effective mass.
@JohnRennie ok.
@JohnRennie what does the last one mean? ^^
What are scattering centres?
10:10
Scattering centres in a semiconductor are places in the lattice where electrons or holes collide with the scattering centre and lose energy. So scattering centres increase the resistance/decrease the conductivity.
@JohnRennie I see. So, correct options are 1, 2 and 3.
Can you explain your reasoning for case (2)?
@JohnRennie very sorry, I misread it as less. 2 is incorrect. More effective mass means more collisions and less conductivity.
More effective mass means the charge carriers accelerate more slowly in response to the electric field.
And option (3)?
@JohnRennie if electric field is more, then conductivity increases as charge carriers are acted upon by a greater force as per $F=qE $.
10:18
@WrichikBasu The current certainly increases with increasing electric field ...
@JohnRennie yes, yes, that's what I meant.
So does the conductivity increase with increasing electric field?
@JohnRennie just a second. I think conductivity is unaffected by electric field. The resistance of a wire remains constant irrespective of current. So, conductivity should also be constant for a material.
@JohnRennie if I'm correct in my above reasoning, then one last thing: is coloumb potential a central potential?
Correct, the conductivity is unaffected by the field. So only (1) is correct.
@WrichikBasu it depends what you mean by the Coulomb potential. The potential due to a single charge is a central potential.
10:36
@JohnRennie ok.
In this, I know that the last two are correct. I'm having problems with the first two.
That's a surprisingly complicated question because at elementary level we use an approximation called the mean field in which the potential is central. But this is just an approximation and the potential is actually not central.
I don't know what answer your question sheet is expecting.
@JohnRennie I see. I'll mark the first as correct and submit. Once I get the answer, I'll tell you want they are expecting.
Anyways, thank you a lot due sparing out time for these questions. :-)
I'll always be grateful to you Sir! @JohnRennie
11:07
@WrichikBasu only the first answer is correct. The potential is not central because it includes the repulsion between electrons and this force isn't directed towards the centre. Angular momentum is only conserved if the potential is spherically symmetric, and since angular momentum isn't conserved there are no $n$ and $l$ quantum numbers.
 
1 hour later…
12:32
@0ßelö7 It certainly defines a physical theory (for integer $p$), but I don't know any actual system that's described by that.
@ACuriousMind I'm interested in noninteger p
Then that's not physical at all, I think
12:56
@ACuriousMind what makes integer $p$ a physical theory?
01:00 - 13:0013:00 - 00:00

« first day (2508 days earlier)      last day (2720 days later) »