POV-Ray : Newsgroups : povray.binaries.images : Granite_21 - the final macro Server Time
28 Mar 2024 17:29:35 EDT (-0400)
  Granite_21 - the final macro (Message 16 to 25 of 30)  
<<< Previous 10 Messages Goto Latest 10 Messages Next 5 Messages >>>
From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 16 Sep 2021 19:00:00
Message: <web.6143cc4fb21f010ccb705ca46e741498@news.povray.org>
"Bald Eagle" <cre### [at] netscapenet> wrote:
> "Samuel B." <stb### [at] hotmailcom> wrote:
> Aren't the HKL values just the vector normals of the planes?
> That's what the wikipedia articles suggested on my brief skim.
>
> Would not plane {y, 0} have an hkl index of <0,1,0>?

That's probably the case with simple cubic crystal systems. Other systems, like
hexagonal, monoclinic, etc., are a bit different. I should try to get a cubic
system to work first.

> Not sure what the overall goal is, so not sure what to suggest to test this.

My goal is to simply convert HKL values to normal vectors, and then to build
crystals from those using intersecting planes.

Sam


Post a reply to this message

From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 16 Sep 2021 19:15:00
Message: <web.6143cf38b21f010ccb705ca46e741498@news.povray.org>
"Bald Eagle" <cre### [at] netscapenet> wrote:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>
> > > Aren't the HKL values just the vector normals of the planes?
> > > That's what the wikipedia articles suggested on my brief skim.
> > >
> > > Would not plane {y, 0} have an hkl index of <0,1,0>?
> > >
> > > Not sure what the overall goal is, so not sure what to suggest to test this.
> > >
> > If I remember correctly, that is indeed the case I believe.
> >
> > --
> > Thomas
>
> So then, to some degree, the unit cell for the crystal should be some form of
> intersection between all of the planes that the hkl indices define. (?)

I think the unit cell defines the dimensions and angles. Attached is a
screenshot from KrystalShaper. The thin-lined, skewed box surrounding the
crystal represents the unit cell. In the lower right area of the image you can
see the numbers "4.913 x 4.913 x 5.4045" which are probably the dimensions of
the cell. The numbers "90 90 120" are the angles. But how to convert these to
proper planes? All my attempts have failed. There is probably a basic solution
and I'm just over-complicating things...

> Does the file format include any data that would serve as an offset?

The data files are in some non-human-readable format. I should probably get a
hex editor and take a look.

> Creating the most basic crystal shapes, probably starting with cubic NaCl, might
> give a clue as to how to code an implementation of that notational system.

Yes, I should start with the most basic thing first.

Sam


Post a reply to this message


Attachments:
Download 'krystalshaper-quartz.png' (18 KB)

Preview of image 'krystalshaper-quartz.png'
krystalshaper-quartz.png


 

From: Bald Eagle
Subject: Re: Granite_21 - the final macro
Date: 17 Sep 2021 06:30:00
Message: <web.61446d55b21f010c1f9dae3025979125@news.povray.org>
"Samuel B." <stb### [at] hotmailcom> wrote:
In the lower right area of the image you can
> see the numbers "4.913 x 4.913 x 5.4045" which are probably the dimensions of
> the cell. The numbers "90 90 120" are the angles. But how to convert these to
> proper planes? All my attempts have failed. There is probably a basic solution
> and I'm just over-complicating things...

I agree that those are the dimensions and angles.
But they are the angles of the unit cell, and ATM I fail to see how you need
them.

What I DO see, is that you have numerical triplets, indicating the hkl indices.
I believe that you can simply use those numbers as the normal vectors for all of
the face planes.  Maybe there's a way to display them with greater precision.
Still not sure about the offset, but maybe once you get the ball rolling, some
light will be shed on that aspect of the data.

Perhaps if there are tables of human readable hkl indices for different crystal
structures, you can compare and contrast and discern a pattern.

And that's all I have before 1st coffee.   :D


Post a reply to this message

From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 18 Sep 2021 17:05:00
Message: <web.6146538cb21f010ccb705ca46e741498@news.povray.org>
"Bald Eagle" <cre### [at] netscapenet> wrote:
> "Samuel B." <stb### [at] hotmailcom> wrote:
> In the lower right area of the image you can
> > see the numbers "4.913 x 4.913 x 5.4045" which are probably the dimensions of
> > the cell. The numbers "90 90 120" are the angles. But how to convert these to
> > proper planes? All my attempts have failed. There is probably a basic solution
> > and I'm just over-complicating things...
>
> I agree that those are the dimensions and angles.
> But they are the angles of the unit cell, and ATM I fail to see how you need
> them.

I think they merely point to the angles and dimensions one would need when
converting HKL indices into proper plane normals. For a cubic system, this would
be straightforward, but for a hexagonal system, something else is needed. I
tried shearing/skewing & scaling cubic coords to produce proper results, but as
said before, I had no luck.

> What I DO see, is that you have numerical triplets, indicating the hkl indices.
> I believe that you can simply use those numbers as the normal vectors for all of
> the face planes.  Maybe there's a way to display them with greater precision.

There is no greater precision. Attached is an image showing the table of HKL
indices of a basic quartz crystal. When you click on any number field, you'll
find it's just an integer. (Except for the distance values, of course.) If I'm
not mistaken, this is because the HKL system only works for the intended
coordinate spaces (aka unit cells).

> Still not sure about the offset, but maybe once you get the ball rolling, some
> light will be shed on that aspect of the data.

Not sure what you mean. The plane offsets? Those are given by the aforementioned
fourth value, "distance."

> And that's all I have before 1st coffee.   :D

Drink 'em if you've got 'em! ;)

Sam


Post a reply to this message


Attachments:
Download 'krystalshaper-quartz-indices.png' (7 KB)

Preview of image 'krystalshaper-quartz-indices.png'
krystalshaper-quartz-indices.png


 

From: Thomas de Groot
Subject: Re: Granite_21 - the final macro
Date: 20 Sep 2021 03:02:01
Message: <614831e9$1@news.povray.org>
Not sure if this might help: http://jcrystal.com/products/winwulff/

I am still thinking how this would help. Fifty five years ago (!) when I 
was introduced to the use of a Wulff net, you were given the model of a 
crystal with the task to translate its features onto the net, 
determining the symmetry planes, the axes of rotation, measuring the 
angles between faces, etc. I suppose the difficulty here is to do the 
opposite, or am I wrong?

-- 
Thomas


Post a reply to this message

From: Bald Eagle
Subject: Re: Granite_21 - the final macro
Date: 21 Sep 2021 15:45:00
Message: <web.614a357db21f010c1f9dae3025979125@news.povray.org>
"Samuel B." <stb### [at] hotmailcom> wrote:
All my attempts have failed. There is probably a basic solution
> > > and I'm just over-complicating things...

Doesn't seem that way.  Having never done any of this type of work, in my
ignorance, I believe I was initially suggesting an overly simplistic approach.

OK, here's the deal.

from: https://en.wikipedia.org/wiki/Miller_index
.... a family of lattice planes is determined by three integers h, k, and
ℓ, the Miller indices. They are written (hkℓ), and denote the family
of planes orthogonal to hb1 + kb2 +lb3, where bi are the basis of the reciprocal
lattice vectors....

from which I can see that the key part is incorporated by reference at
https://en.wikipedia.org/wiki/Reciprocal_lattice

The reciprocal lattice is the set of all vectors Gm, that are wavevectors of
plane waves in the Fourier series of a spatial function which periodicity is the
same as that of a direct lattice Rn

So:

The guy you want to talk to is Michael Joseph Waters - a postdoctoral researcher
in X-ray crystallography at Northwestern.  Nice guy - we emailed back and forth
a while back.
"NiO is a odd material, it's antiferromagnetic so there are alternating planes
of spin up and spin down electrons. If you learn Miller indices, these are on
(111) planes."
Also:
https://ciderware.blogspot.com/2016/08/getting-isosurfaces-from-data-grid-into.html

Find his contact info here:
https://mtd.mccormick.northwestern.edu/group/

He's likely a busy guy, but he may have some advice on how to approach
unraveling this all.  I'm guessing it might take a 3rd party software package -
or maybe not, if we can use the FFT that I coded up a while back.


Post a reply to this message

From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 21 Sep 2021 19:15:00
Message: <web.614a66a9b21f010ccb705ca46e741498@news.povray.org>
Thomas de Groot <tho### [at] degrootorg> wrote:
> Not sure if this might help: http://jcrystal.com/products/winwulff/
>
> I am still thinking how this would help. Fifty five years ago (!) when I
> was introduced to the use of a Wulff net, you were given the model of a
> crystal with the task to translate its features onto the net,
> determining the symmetry planes, the axes of rotation, measuring the
> angles between faces, etc. I suppose the difficulty here is to do the
> opposite, or am I wrong?

That might help, but I'm not sure. Ideally, I'd like to perform all the
necessary conversions in POV-Ray. I'll check that out and see if it shines a
light on the subject.

Sam


Post a reply to this message

From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 21 Sep 2021 19:40:00
Message: <web.614a6c5fb21f010ccb705ca46e741498@news.povray.org>
"Bald Eagle" <cre### [at] netscapenet> wrote:
> "Samuel B." <stb### [at] hotmailcom> wrote:
> > > > All my attempts have failed. There is probably a basic solution
> > > > and I'm just over-complicating things...
>
> Doesn't seem that way.  Having never done any of this type of work, in my
> ignorance, I believe I was initially suggesting an overly simplistic approach.
>
> OK, here's the deal.
>
> from: https://en.wikipedia.org/wiki/Miller_index
> .... a family of lattice planes is determined by three integers h, k, and
> ℓ, the Miller indices. They are written (hkℓ), and denote the family
> of planes orthogonal to hb1 + kb2 +lb3, where bi are the basis of the reciprocal
> lattice vectors....
>
> from which I can see that the key part is incorporated by reference at
> https://en.wikipedia.org/wiki/Reciprocal_lattice
>
> The reciprocal lattice is the set of all vectors Gm, that are wavevectors of
> plane waves in the Fourier series of a spatial function which periodicity is the
> same as that of a direct lattice Rn

Thank for the time you took to research this! It's indeed probable that the
reciprocal lattice is the way forward. I'll read through the articles and try to
glean some useful info from them.

Unfortunately, I've got two things holding me back: my inability to parse
mathematical formulas due to only being self-taught; and my inconsistent mental
energy. Most people suffer from the latter, but it seems I am particularly
affected by such fluctuations :S

> So:
>
> The guy you want to talk to is Michael Joseph Waters - a postdoctoral researcher
> in X-ray crystallography at Northwestern.  Nice guy - we emailed back and forth
> a while back.
> "NiO is a odd material, it's antiferromagnetic so there are alternating planes
> of spin up and spin down electrons. If you learn Miller indices, these are on
> (111) planes."

I'm intrigued by that. What sorts of special properties does Nickle(II) oxide
possess, I wonder?

> Also:
> https://ciderware.blogspot.com/2016/08/getting-isosurfaces-from-data-grid-into.html
>
> Find his contact info here:
> https://mtd.mccormick.northwestern.edu/group/
>
> He's likely a busy guy, but he may have some advice on how to approach
> unraveling this all.  I'm guessing it might take a 3rd party software package -
> or maybe not, if we can use the FFT that I coded up a while back.

FFT=fast Fourier transform? What have you been using it for? (Sorry for being
out of the loop, haha.)

And I'd hate to bother him to pick his brain, but I do know some people like to
talk. Heck, maybe the creator of KrystalShaper would be willing to share the
code he used to convert HKL indices to planes.

Thanks again, and I will be looking into the matter further. I'm betting the
conversions for various crystal symmetries are actually super easy to
implement...

Sam


Post a reply to this message

From: Bald Eagle
Subject: Re: Granite_21 - the final macro
Date: 21 Sep 2021 22:00:00
Message: <web.614a8d3bb21f010c1f9dae3025979125@news.povray.org>
"Samuel B." <stb### [at] hotmailcom> wrote:

> Unfortunately, I've got two things holding me back: my inability to parse
> mathematical formulas due to only being self-taught; and my inconsistent mental
> energy. Most people suffer from the latter, but it seems I am particularly
> affected by such fluctuations :S

I'm mostly self-taught as well.  Let me know what you need to know, and I'm
always happy to try to explain.

> I'm intrigued by that. What sorts of special properties does Nickle(II) oxide
> possess, I wonder?

"Exploring the interplay between electronic doping, magnetism, and Ni diffusion
in NiO from first principles."

NiO and NiO solid solutions are typically the first passivating oxides to form
during the corrosion of Ni-rich alloys. Ni diffusion from the alloy interface,
through the oxide, to the exposed surface is the rate-limiting kinetic process
for continued oxidation. From first principles, we calculate the effects of
magnetic order and electronic doping on the minimum energy diffusion pathways.
We elucidate the origin of the diffusion barrier dependencies by assessing
changes in the defect states and chemical bonding along the diffusion paths. Our
work highlights the complexity of diffusion in even structural simply but
strongly correlated oxides with practical recommendations for corrosion
resistance that may be transferable to diffusion-mediated redox processes in
energy storage materials.

> FFT=fast Fourier transform? What have you been using it for? (Sorry for being
> out of the loop, haha.)

http://news.povray.org/povray.binaries.images/thread/%3C58bfe702%241%40news.povray.org%3E/
http://news.povray.org/povray.binaries.images/thread/%3Cweb.58bd5ea376ca50b7c437ac910%40news.povray.org%3E/?ttop=435711
&toff=550

I haven't had much of a chance to really use it for anything.
It was suggested that I might be able to use it for my gear tooth counting
project:

http://news.povray.org/povray.general/thread/%3C59371017%241%40news.povray.org%3E/
http://news.povray.org/povray.binaries.images/thread/%3C593bd19f%40news.povray.org%3E/

> And I'd hate to bother him to pick his brain, but I do know some people like to
> talk. Heck, maybe the creator of KrystalShaper would be willing to share the
> code he used to convert HKL indices to planes.

They are supported by millions of our tax dollars.  If he wants to answer your
email, he will.  If he doesn't, then you made a reasonable request, and you can
try elsewhere.  And really, he seemed very cool and happy to talk about his
research, and explain some of the basics that I was ignorant of.  Politely
asking for some expert assistance with one thing isn't really "bothering."

> Thanks again, and I will be looking into the matter further. I'm betting the
> conversions for various crystal symmetries are actually super easy to
> implement...

Probably.  Hopefully we'll find out.


Post a reply to this message

From: Samuel B 
Subject: Re: Granite_21 - the final macro
Date: 23 Sep 2021 19:45:00
Message: <web.614d1063b21f010ccb705ca46e741498@news.povray.org>
"Bald Eagle" <cre### [at] netscapenet> wrote:
> "Samuel B." <stb### [at] hotmailcom> wrote:
>
> > (...) my inability to parse mathematical formulas due to only being self-taught
(...)
>
> I'm mostly self-taught as well.  Let me know what you need to know, and I'm
> always happy to try to explain.

I might just take you up on that. I know the rules of PEMDAS, and I know
functions and their usages in programming languages, but when I see formulas on
Wikipedia or elsewhere I see many symbols representing unfamiliar equations. I
know it's only a matter of searching and finding, but I'm certain these things
are more easily grokked by the formally-trained.

> > I'm intrigued by that. What sorts of special properties does Nickle(II) oxide
> > possess, I wonder?
>
> "Exploring the interplay between electronic doping, magnetism, and Ni diffusion
> in NiO from first principles."
>
> NiO and NiO solid solutions are typically the first passivating oxides to form
> during the corrosion of Ni-rich alloys. Ni diffusion from the alloy interface,
> through the oxide, to the exposed surface is the rate-limiting kinetic process
> for continued oxidation. From first principles, we calculate the effects of
> magnetic order and electronic doping on the minimum energy diffusion pathways.
> We elucidate the origin of the diffusion barrier dependencies by assessing
> changes in the defect states and chemical bonding along the diffusion paths. Our
> work highlights the complexity of diffusion in even structural simply but
> strongly correlated oxides with practical recommendations for corrosion
> resistance that may be transferable to diffusion-mediated redox processes in
> energy storage materials.

Not sure what this means, but am I jumping to conclusions by thinking this means
NiO-rich alloys corrode more easily from the 'sides'? Similar to how a
stratified bedrock can be more easily penetrated by weathering forces attacking
between its layers?

> > FFT=fast Fourier transform? What have you been using it for? (Sorry for being
> > out of the loop, haha.)
>
>
http://news.povray.org/povray.binaries.images/thread/%3C58bfe702%241%40news.povray.org%3E/
>
http://news.povray.org/povray.binaries.images/thread/%3Cweb.58bd5ea376ca50b7c437ac910%40news.povray.org%3E/?ttop=4357
11
> &toff=550
>
> I haven't had much of a chance to really use it for anything.
> It was suggested that I might be able to use it for my gear tooth counting
> project:
>
> http://news.povray.org/povray.general/thread/%3C59371017%241%40news.povray.org%3E/
>
http://news.povray.org/povray.binaries.images/thread/%3C593bd19f%40news.povray.org%3E/

Very interesting!

After reading that I was trying to figure out a better way to count gear teeth,
but was drawing a blank. Somebody in one of those threads mentioned (to heavily
paraphrase) performing a polar-to-cartesian transformation on an image of a gear
and evaluating that (probably by tracing rays horizontally and incrementing a
counter). I wonder how that would work out?

And I think you mentioned possibly using an FFT to approximate 3D geometry using
isosurfaces. I do believe there is some potential there.

> > And I'd hate to bother him to pick his brain, but I do know some people like to
> > talk. Heck, maybe the creator of KrystalShaper would be willing to share the
> > code he used to convert HKL indices to planes.
>
> They are supported by millions of our tax dollars.  If he wants to answer your
> email, he will.  If he doesn't, then you made a reasonable request, and you can
> try elsewhere.  And really, he seemed very cool and happy to talk about his
> research, and explain some of the basics that I was ignorant of.  Politely
> asking for some expert assistance with one thing isn't really "bothering."

I heard that many researchers are willing to share their research. In fact, it's
been said if you can't find a paper for free on sci-hub, then maybe you should
contact the author(s) directly. I love the fact that people are willing to share
their findings, regardless of profit.

Sam


Post a reply to this message

<<< Previous 10 Messages Goto Latest 10 Messages Next 5 Messages >>>

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.