|
Science Forum Index » Physics - Relativity Forum » fluid dynamics made simple
Page 1 of 1
|
| Author |
Message |
| Thomas Heger |
 Posted: Wed Apr 30, 2008 4:20 am |
|
|
|
Guest
|
Hi Ng
I would like to show you a scheme to calculate all kinds of fluids. It
could cover all types of things that could possibly flow: gas, solids,
fluids, charges...
It goes like this:
You setup a sheet of cells. You twist those cell in certain directions
and give each cell an amplitude. Each cell gets a number.
According to some simple rules you perform timelike steps. This is like
piling up those sheets. You don't need to care for things like heat,
viscosity or friction. Those units are covered within. The trick is the
right setup. But you can even model very strange fluids or electric
currents with that.
That sheet is representing a kind of tube, but you could even model
other shapes if you have enough cpu.power and time (of cause you would
use a computer for this).
Those cells could be small, but it's not neccesary. You could make
things like a drop.
That picture is like a energy flow passing perpendicular to that sheet.
To get back your flow, you have to follow each cell passing through this
stack of sheets. Than you can follow their trajectory and get your flow
modelled.
The art is where to twist the cells and how to setup the amplitudes. I
named it 'the cube' and describe it here:
http://docs.google.com/Presentation?id=dd8jz2tx_3gfzvqgd6
Thomas Heger |
|
|
| Back to top |
|
| N:dlzc D:aol T:com (dlzc) |
| Posted: Wed Apr 30, 2008 8:50 pm |
|
|
|
Guest
|
Dear Thoma Heger:
"Thomas Heger" <tomheg_nospamxxx@web.de> wrote in message
news:fv9dl9$h02$03$1@news.t-online.com...
....
Quote: I would like to show you a scheme to calculate
all kinds of fluids.
Congratulations, you have just reinvented "computational fluid
dynamics".
Been around almost as long as computers.
David A. Smith |
|
|
| Back to top |
|
| Thomas Heger |
| Posted: Wed Apr 30, 2008 9:21 pm |
|
|
|
Guest
|
"N:dlzc D:aol T:com (dlzc)" <dlzc1@cox.net> schrieb im Newsbeitrag
news:Kl9Sj.7510$DG.3092@newsfe10.phx...
Quote: Dear Thoma Heger:
"Thomas Heger" <tomheg_nospamxxx@web.de> wrote in message
news:fv9dl9$h02$03$1@news.t-online.com...
...
I would like to show you a scheme to calculate
all kinds of fluids.
Congratulations, you have just reinvented "computational fluid dynamics".
Been around almost as long as computers.
David A. Smith
Sorry, I didn't know this. Never heard of that bevor. My idea was something
I thought could be a usefull application of my 'simple theory'.
thomas Heger |
|
|
| Back to top |
|
| Thomas Heger |
| Posted: Wed Apr 30, 2008 9:42 pm |
|
|
|
Guest
|
"N:dlzc D:aol T:com (dlzc)" <dlzc1@cox.net> schrieb im Newsbeitrag
news:Kl9Sj.7510$DG.3092@newsfe10.phx...
Quote: Dear Thoma Heger:
"Thomas Heger" <tomheg_nospamxxx@web.de> wrote in message
news:fv9dl9$h02$03$1@news.t-online.com...
...
I would like to show you a scheme to calculate
all kinds of fluids.
Congratulations, you have just reinvented "computational fluid dynamics".
Been around almost as long as computers.
David A. Smith
That approach is quite different to mine. my idea uses general relativity in
quaternial form to calculate fluids. It could be *any* kind of fluid: gas,
charge, solid, liquids, glass, geologic formations ...
And I only use dimensionless numbers in two dimensions (for tubes) as setup.
And it's supposed to be of infinite accuracy, depending only on the setup
and the cpu-time.
Thomas Heger |
|
|
| Back to top |
|
| N:dlzc D:aol T:com (dlzc) |
| Posted: Wed Apr 30, 2008 10:08 pm |
|
|
|
Guest
|
Dear Thomas Heger:
"Thomas Heger" <hballo@hotmail.com> wrote in message
news:fvbanl$j9b$00$1@news.t-online.com...
....
Quote: That approach is quite different to mine. my idea
uses general relativity in quaternial form to
calculate fluids.
Then it does not work. Check out "Navier Stokes".
And your methodology is still a finite element method, and is
quite old.
David A. Smith |
|
|
| Back to top |
|
| Thomas Heger |
| Posted: Thu May 01, 2008 2:38 am |
|
|
|
Guest
|
"N:dlzc D:aol T:com (dlzc)" <dlzc1@cox.net> schrieb im Newsbeitrag
news:AuaSj.107209$497.96682@newsfe14.phx...
Quote: Dear Thomas Heger:
"Thomas Heger" <hballo@hotmail.com> wrote in message
news:fvbanl$j9b$00$1@news.t-online.com...
...
That approach is quite different to mine. my idea
uses general relativity in quaternial form to
calculate fluids.
Then it does not work. Check out "Navier Stokes".
And your methodology is still a finite element method, and is quite old.
David A. Smith
right now I've found a trick, that is simply shocking. If you think about
quaternions like I do, you could rebuild all of quantum physics out of a few
simple principles. Its really simple. And it will be a shock to anybody that
will understand this. I named this "the cube". With this relation you can
twist matter into light (and back). If you encapsulate relation to an
operator and object to another one you can do all kinds of things with only
this.
Thomas Heger |
|
|
| Back to top |
|
| N:dlzc D:aol T:com (dlzc) |
| Posted: Thu May 01, 2008 8:07 am |
|
|
|
Guest
|
Dear Thomas Heger:
"Thomas Heger" <hballo@hotmail.com> wrote in message
news:fvbs2r$228$02$1@news.t-online.com...
....
Quote: right now I've found a trick, that is simply shocking.
If you think about quaternions like I do,
You are turning into a crank, Thomas. You have not invented the
greatest thing since sliced bread. If you have, please generate
a closed form solution to either EFE or Navier-Stokes. The
scientific community will beat a path to your door.
Its nice that you are pleased with it, but as has been pointed
out, it has its limitations. Your new playground has some old
toys scattered around it, if you'll notice.
David A. Smith |
|
|
| Back to top |
|
| dlzc |
| Posted: Thu May 01, 2008 11:51 am |
|
|
|
Guest
|
Dear Thomas Heger:
On May 1, 2:24 pm, "Thomas Heger" <hba...@hotmail.com> wrote:
Quote: "N:dlzcD:aol T:com (dlzc)" <dl...@cox.net> schrieb im Newsbeitragnews:1gjSj.1105$152.783@newsfe12.phx...
...
Its nice that you are pleased with it, but as has been
pointed out, it has its limitations. Your new
playground has some old toys scattered around it, if
you'll notice.
...
That my suggested algorithm. you think something like
this is used?
Yes.
I'd recommend a course in "finite element analysis", if you want to
know more. Or at least "numerical methods". The pictures running
through your head ran through the heads of Newton, Euler, and many
more. Nothing to be ashamed of, but time to see the rest of the
*current* landscape.
To get your feet wet, go here:
http://arxiv.org/find
... search in "gr-qc" and search for "numerical". Some of the hits
will be apropriate... just barely over your head, or at least briefly
so, so that you can see the direction they are heading. And there
will be "cognates" ("keywords") that you can use to do further
searches.
David A. Smith |
|
|
| Back to top |
|
| Thomas Heger |
| Posted: Thu May 01, 2008 4:24 pm |
|
|
|
Guest
|
"N:dlzc D:aol T:com (dlzc)" <dlzc1@cox.net> schrieb im Newsbeitrag
news:1gjSj.1105$152.783@newsfe12.phx...
Quote: Dear Thomas Heger:
"Thomas Heger" <hballo@hotmail.com> wrote in message
news:fvbs2r$228$02$1@news.t-online.com...
...
right now I've found a trick, that is simply shocking.
If you think about quaternions like I do,
You are turning into a crank, Thomas. You have not invented the greatest
thing since sliced bread. If you have, please generate a closed form
solution to either EFE or Navier-Stokes. The scientific community will
beat a path to your door.
Its nice that you are pleased with it, but as has been pointed out, it has
its limitations. Your new playground has some old toys scattered around
it, if you'll notice.
I guess there is a danger to get nuts about ideas and trying to solve the
riddles of the world. So I guess it's a good advice to spent more time away
from the computer. I guess i'm finished with my idea anyhow, but let me
explain the idea of that algorithm. There might be someone interested, since
fluids are very importent in all kinds of productions.
So it goes like this. Its base an quaternions rotation and the formular is
this: w'=q*w*q^-1.that is for the neighbours. In direction of a rotation it
is w'=q*w. The neighbours get twisted. But that twist is then pointing
sydeways and has to use the second formular and now the neighbours are in
the direction that was the future before and so on. So the influences go in
a zig-zag path and then return zig-zag back to the origin, but leave out one
step. By this process a timestep is performed. All amplitudes are summend on
the path in amount and direction. It has to be done in all directions and
for all cells. There might be a cutof. And there might be bigger cells. The
size is not that important. After any timestep you keep the results and redo
the process. The patterns within are tracked by a pattern recogniton
programm and followed over the timelike evoloution.
That my suggested algorithm. you think something like this is used? |
|
|
| Back to top |
|
| |
