« first day (3227 days earlier)      last day (1695 days later) » 

12:46 AM
I'm reading von Neumann's book on QM and I'm slightly confused by a simple point. He writes, "The energy is a given function of the coordinates and their time derivatives: $E = L(q_1,..,q_k;\dot{q}_1,...,\dot{q}_k) = H(q_1,...,q_k;p_1,...,p_k)$. (This $H$ is the Hamiltonian function.)
Does $L(....)$ denote the Lagrangian? Why does the second equality hold?
 
1:09 AM
Usually it denotes the Lagrangian...and usually H=qp-L...not sure why it's in that form in that equation...
 
 
7 hours later…
It looks as if the TiO2 particles have flocculated (i.e. stuck together)
The suspension should look like milk. You shouldn't be able to see individual particles.
 
I haven't changed the H2O NaOH mixture for about a month and a half, should I make a new one or that's not the problem? I have another capsule where it looks like milk, I'll try with it.
 
@bolbteppa about what?
 
Also when I apply the electric field on the milk-like mixture, the I should see a clear distinction between the TiO2 particles and the water (they should split)?
 
 
1 hour later…
9:35 AM
Visual Studio 2019 now supports Python, and it seems to work very well. I've only had a quick play so far, but everything I've tried works fine including packages like numpy that aren't included in a default Python install.
 
9:46 AM
I'm still waiting APL in Visual Studio
 
Some messing around further and I'm really impressed by the Python support in VS2019.
Visual Studio is by far the best IDE for C++, which is normally all I use it for, and all the good stuff works in Python too.
 
I mostly use it for C#
C++ I don't bother and just use gpp
I just compile it by hand by translating it into machine code
 
10:01 AM
I toggle in the machine code in binary using an array of switches.
(I have actually done this back in the early 1980s)
 
 
3 hours later…
12:49 PM
Need to get this low energy effective action for superstrings thing down
 
1:04 PM
its GR
GR and also fucknig dilaton field
Tho dilaton field is good
u can get wormholes without breaking the NEC
 
Yo yo
I know Hamilton now
Can talk fancy stuff
 
Do u know about constraints
and the Legendre bundle
 
@AvnishKabaj I heard it has good reviews
 
@AvnishKabaj do you know that $\frac{\partial S}{\partial t} = - H$
 
1:22 PM
One thing I would actually like to know is why there is that general notion that you can only do Hamiltonian mechanics on a globally hyperbolic spacetime
You can define Hamiltonians on arbitrary spacetimes
I'm guessing you can't have unique hamiltonian evolutions, though
But nobody ever bothers to prove it
I have even see people defining Hamiltonians on non-globally hyperbolic spacetimes bc why not
Carlip does it for 2+1D Schwarzschild
 
@bolbteppa yos
Partial derivatives as well
@SirCumference :)
(:
:)
 
So $S = \int L dt$ is the action, and since $L = p \frac{dq}{dt} - H$ is the Legendre transform of $L$, we see $H = p \frac{dq}{dt} - L$, so that $S = \int L dt = \int [p \frac{dq}{dt} - H]dt = \int (p dq - H dt)$. From this you can see $\frac{\partial S}{\partial t} = - H$ easily.
Now the Schrodinger equation in quantum mechanics comes from the fact that a wave function $\Psi$ should reduce to a function of the form $\Psi \approx e^{i\frac{S}{\hbar}}$ in the classical limit, where $\hbar$ is a quantity that makes the argument of the exponential function dimensionless, and so taking the time derivative
$\frac{\partial \Psi}{\partial t} \approx \frac{\partial }{\partial t} e^{i\frac{S}{\hbar}} = i\frac{1}{\hbar} \frac{\partial S}{\partial t} e^{i\frac{S}{\hbar}} = - \frac{i}{\hbar} H \Psi $ we find the Schrodinger equation $ i \hbar \frac{\partial \Psi}{\partial t} = H \Psi$
Now you know some quantum too
 
Niiiiiiiiiiiice
 
The thing I like to show that Hamiltonian mechanics makes sense is to introduce a bunch of extra variables
Like instead of $S[x] = \int L(x, \dot{x}) dt$, have $S[x, v] = \int L(x, v) dt$
But to have the proper relation between $v$ and $x$, you need some Lagrange multiplier $p(v - \dot{x})$
 
1:37 PM
Legendre transform is predicated on $\dot{x} = v$ right
 
Yeah
But if you're doing Hamiltonian constraints, it's nice to have that form
Well, no
The Legendre transform is predicated upon having $v = v(p)$
By definition $\dot{x} = v$ on shell
the hard part is inverting $p(v)$
Then your action is just $$S[x, v, p] = \int \left[ L - pv \right] + p \dot{x}$$
And you define $H = L - pv$
And the EOM imply $$\dot{p} = \frac{\partial L}{\partial q}$$
And if you're lucky you can use this to find some homeomorphism between $p$ and $v$
It's a nice little trick which is both easier than Legendre transform and also you can use it for constrained Hamiltonians
Errr the important part is more $$p = \frac{\partial L}{\partial v}$$
But if you can't invert it then you still have your Hamiltonian making sense that way
 
1:56 PM
This seems like something that can just be ignored :p
 
It can, but I think it's a neater demonstration that way
Then you can split velocities between the invertible and non-invertible ones
 
I don't see why it's better, think messing with the idea that $v = dq/dt$ throws everything out the window
 
and the non-invertible ones become the Lagrange multipliers for constraints
 
It's already hard enough to accept that $p \neq \gamma mv$
 
heh
Part of becoming a physicist is accepting that quantities are as they are defined :p
Momentum is what momentum is defined as in the theory
Be it $p = mv$, $p = \partial L / \partial q$, $p = T_{0i}$, $p$ is the translation generator or whatever
and hopefully you'll find out how they all relate
Obviously $p = p_2(\psi(y))$, $y \in T^*M$
$p_2$ the projection operator of the second object of an ordered pair
and $\psi$ the local trivialization
 
2:34 PM
the true wisdom u must learn to become a theoretical physicist is that all theoretical entities are equally as real
 
Except for the complex stuff :p
Like the spinors which have no Majorana's in certain dimensions
 
slap
 
How can she slap!?
 
3
Q: Under what conditions is a vector-spinor gamma trace free

NeuneckI came upon the concept of irreducible vector-spinors while trying to simplify an expression involving the gravitino field. It is claimed that an irredicible vector-spinor is gamma-traceless, i.e. $$ \Gamma^M \psi_{\alpha M} = 0.$$ Are there conditions on the irreducibility besides the above eq...

1
Q: Gamma traceless

Filip GeorgijevskiI read this Under what conditions is a vector-spinor gamma trace free. And also read many papers about higher spin, but no one explains why irreducible spinor is gamma traceless spinor? Can anyone explain this?

Absolutely amazing
 
3:35 PM
@JohnRennie I managed to get it to look like milk, but it still doesn't work.
 
To solve Schrodinger's equation for the hydrogen atom (time-independent), $$ \dfrac{-\hbar^2}{2m} \nabla_x^2 \psi(x) + v(x)\psi(x) = E \psi(x) $$, How to convert the nabla from the cartesian coordinate to the Spherical ones?
This shit gives me issues.
 
@AbhasKumarSinha I'm not even smart enough to have those issues :D
 
@NovaliumCompany wym?
 
@AbhasKumarSinha I'm not smart enough to understand those complicated equations and therefore I cannot have those issues. (of having problems with them (the equations))
 
@NovaliumCompany lol, honestly, it's very easy...
 
3:39 PM
@AbhasKumarSinha maybe it is, but I haven't spend time learning it
I barely managed to get some sense of calculus
 
I too thought that equation to be very complicated once, until, I saw how it works
 
@bolbteppa :O
 
@bolbteppa sir, is it divergence operator? $\nabla$, I've seen multiple names for it.
 
Yes
There are faster ways but one should learn these more direct ways first
 
3:42 PM
@bolbteppa Faster means? formulas to solve the equation or something other?
 
Is Curl and divergence same?
 
No
 
Oh okay, one is scalar and one is vector
 
@JohnRennie Maybe my electric field is not strong enough to beat gravity? I'm putting about 320V across 0.5 cm
@JohnRennie Is it possible that I've put too much TiO2 particles?
 
3:54 PM
@bolbteppa Then, how to solve the equation (time independent) for hydrogen atom, (completed the conversion part)
 
separation of variables
 
oh okay, I'm using WolframAlpha for a moment
 
@NovaliumCompany How come you decided to use water instead of oil? The Wikipedia article says that electrophoretic e-paper uses oil.
 
@PM2Ring hey!! :)
 
@PM2Ring JohnRennie recommended I try with water first, see if it works, and then move to oil and so on. I tried with baby oil, but I can't charge the particles. If I put a little water in the oil with TiO2 particles, a bubble of water and TiO2 forms. The drying method doesn't work since removing the water removes the charge.
 
4:10 PM
Oh, ok. Did you try adding some detergent to the water + TiO2 before mixing it with the baby oil?
 
My guess is that your current setup with water doesn't work because water is conductive.
 
@PM2Ring What can I try to do?
 
@NovaliumCompany what you are doing?
 
@AbhasKumarSinha long story :P
 
4:14 PM
hahaha Oh okay
 
Aug 18 at 14:50, by PM 2Ring
You also need a surfactant. A little bit of detergent added to the TiO2 sludge while it's evaporating will probably work. John Rennie may have better suggestions.
@NovaliumCompany Hopefully, a little bit of water will be ok. That way, you have your charged particles, some surfactant, and a liquid that's mostly non-conductive.
As I said a few weeks ago, your cell is basically a capacitor, so you want an insulator between a pair of conductive plates.
 
So here's what I can try: I put a solution of H2O, NaOH and TiO2 particles to evaporate on a plate while I put some soap. I don't want the whole water to evaporate so I leave a bit. I end up with a sludge. I put that sludge in baby oil.
 
baby oil well, this is getting more funnier
 
@NovaliumCompany Yes. I think liquid detergent will be better than solid soap. But that's just a hunch.
 
@PM2Ring What about the liquid type of soap?
 
4:22 PM
@NovaliumCompany Sure.
@AbhasKumarSinha He's trying to make a giant e-paper pixel. en.wikipedia.org/wiki/Electronic_paper#Electrophoretic
 
@PM2Ring what can I do with it?
 
@AbhasKumarSinha It's the material that Amazon Kindle screens are made from and stuff. It's "electronic paper". You can think of it kinda like paper where what is written on it is electronically controlled, or like a screen that doesn't use any backlighting, and only changes the colour of reflected light
 
Traditional soap is a sodium (or potassium) salt of a fatty acid. That makes one end of the molecule oil-soluble, and the other end water-soluble. But I'm not sure if that sodium will be a good thing or a bad thing in the e-paper cell.
 
@JMac 'electronically controlled' can't be made at home, if I'm right?
if it can be, someone send me the screen, I'll program the microcontroller for you :)
 
@AbhasKumarSinha He doesn't need a microcontroller, he's just making a single giant 1 cm diameter pixel.
 
4:30 PM
@AbhasKumarSinha Well, it probably can, and I'm pretty sure that's what Novalium is working towards. Step one is getting a single "unit" which he is able to flip between two colours using electricity. I get the feeling that Novalium is trying to do it for themselves anyways, as a learning experience.
 
@NovaliumCompany If your screen works, I'll program microcontrollers for you :)
@JMac seems veryyy interestingggggggggggggggggg
 
@AbhasKumarSinha Thanks :D
 
I want to see more of it
 
@PM2Ring Should I add the liquid soap at the beginning of the evaporation?
 
@NovaliumCompany That will slow down the evaporation, so I'd end it towards the end.
 
4:34 PM
@NovaliumCompany What will it show at first? I mean, what are you trying to make the screen show? Are you using a computer to control it or what?
 
@AbhasKumarSinha I can't reveal much detail for now sorry :P
@PM2Ring oks
 
@AbhasKumarSinha It's just a single pixel, not a screen! Think of it as the e-paper equivalent of a single LED.
 
@NovaliumCompany oh okay XD
@PM2Ring single pixels? Can it be extended to more? Does it works in the same way, in which classical calculator screen works?
 
@AbhasKumarSinha I think the point is that it's difficult to extend it to more before you even have one working. Typical e-paper is basically a normal screen; but making it from scratch with your own plans would require you to make sure you can get a single pixel to work before trying a screen with multiple
 
Jul 25 at 13:44, by Novalium Company
Why aren't there consumer available e-paper capsules? Just like there are individual LEDs instead of LED displays (or LCD).
 
4:41 PM
oh okay
 
Finally! :)
 
I'm watching The Godfather. 10 mins in and I'm hooked
 
@NovaliumCompany nice movie
 
It's certainly a great, powerful work. But I don't think "nice" really applies to it...
 
#Chandrayaan2 is now on it's mission to the moon, will be touching the lunar surface tomorrow at 1:30 AM IST, I tell you, if it gets successful, then I won't sleep till 2:30 tomorrow.
 
5:03 PM
can two particles accelerate away from each other because they repel each other, s.t. in a finite amount of time the distance between them is infinite
 
I just realized the mafia guy from that Zootopia movie is entirely based on The Godfather
 
@Ultradark "infinite" is not an allowed value for "distance".
 
@ACuriousMind then any distance besides infinity
in other words the time vs distance plot is asymptotically bounded at let's say $t=1$
as t approaches $t=1$ the distance approaches infinity
 
As distance is the second integral of acceleration and $F=ma$, the answer is no because there are no forces that increase with distance and especially none that diverge for any non-zero distances.
 
okay so it's not physically possible
how is the universe accelerating apart?
 
5:16 PM
Ah, I see where you're coming from. The poorly understood reason is called "dark energy", and it involves general relativity where we shouldn't apply Newtonian physics as there isn't really any "force" to the expansion. See physics.stackexchange.com/q/2110/50583 and its answers for some discussions of the universe's expansion.
 
how poorly understood is dark energy?
 
Also, hyperphysics.phy-astr.gsu.edu/hbase/Astro/hubble.html for a brief summary, with equations & diagrams.
 
so it's more of a global acceleration
at the local level things are not accelerating away from each other
 
5:30 PM
@Ultradark Well, the expansion is mostly linear, with a small acceleration component, which can often be ignored, unless you're looking at truly vast distances & time scales. And even when you don"t ignore the acceleration, the expansion rate is so small that gravity overwhelms it, until you get to the scale of galaxy clusters. Individual galaxies within a cluster aren't expanding away from each other because they are bound by gravity.
 
5:53 PM
@Semiclassical
Block[{r = 10},
 Plot[
  HypergeometricU[-e, 1, r^2]/(Gamma[1 + e] Exp[r^2/2])
  , {e, 0, 60}
  ]
 ]
looks a hell of a lot like an Airy function
(which would definitely help simplify things)
I can't find a clean connection, though
though then again, setting r=0.1 kinda breaks that
0
Q: Is there a clean relation between the Tricomi function $U(-a,b,z)$ (as a function of its first argument) and an Airy function?

E.P.Following a suggestion from this thread on Computational Science, I had a brief look at a plot of $$ f(E) = \frac{U(-E,1,R)}{\Gamma(1+E)}, $$ i.e. the Tricomi confluent hypergeometric function as a function of its first parameter, which looks like this for $R=10$. I'm struck by the resemblanc...

 
6:24 PM
@EmilioPisanty I see what you mean, yeah. I’m not yet seeing an easy relation tho
It may be that, within that regime, a similar approximation applies to both
Which would not be so terribly strange: the Airy and Tricomi functions are both obtained as particular confluences of the hypergeometric function
So it may be that they have a nice relation in that regime even tho they don’t everywhere
On those grounds, it may be worth comparing their contour integral representations since that’s what I’d subject (after sensibly deforming the contour) with Laplace’s method
That said, there is one obvious difference between them: You’re plotting U/Gamma as a function of its parameter, whereas Ai is plotted with respect to its argument (since it has no parameters)
So that may complicate any analogy between them in this regime
 
6:47 PM
I just ran into something interesting. Humans can live near a black hole. The dead zone in the 80'S Sagittarius A has a dead zone of 3,200 light years. No wonder why black holes are only known as deadly. But a recent experiment has proven quite the opposite. According to Harvard University Manaswi Lingam the dead zone is much smaller than originally thought at 140 light years.
At 140 light years from the black hole Sagitarious A life can begin and flourish.
At that distance the habitable zone begins. At this distance planets the mass of the earths will have there atmospheres. The radiation at this distance isn't so strong as to destroy all living things. The ultraviolet radiation at that distance can breakdown molecules to produce proteins, lipids and DNA. The radiation can also stimulate photosynthesis. Which is the most important processes for oxygen.
 
Fear thee
Fear thee even more
 
1 Million earth mass planets can be in the habitable zone. This is according to the astrophysicists at the Bordeaux observatory. They can be placed on 400 rings with 2,500 planets on each. The distance from each planet will be 10 times less than between the moon and the earth.
 
 
2 hours later…
8:54 PM
@bolbteppa I am fearful :)
 
rob
9:14 PM
@ScientistSmithYT Probably not
 
@bolbteppa :)
@rob But a black hole doesn't have the same mass as a normal star.
 
rob
@ScientistSmithYT The problem is interactions between orbiting bodies.
 
@rob Huh... Maybe the physicist didn't take that into account. But why wouldn't he have? He must have taken that in to account.
 
rob
@ScientistSmithYT Do you have a link to the article?
 
@rob Not at the moment, but I can get it. I hate timing though. I have to go do some stuff. But I'll get that paper for you at least by tonight. I have plans to meet up with some students of mine. I'm teaching a HV electricians course. I'm super young, but I have a lot of skills. I am of course going to be supervised by a few other HV Electricians.
 
rob
9:28 PM
@ScientistSmithYT I find this article about a BH Goldilocks zone, but nothing about a million co-orbiting planets.
@ScientistSmithYT Oh, no rush. One of the nice things about SE chat is that the notification system lets you leave a conversation and come back later.
Good luck with your class. Teaching is exciting!
 

« first day (3227 days earlier)      last day (1695 days later) »