<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
    <channel>
        <title>LBMethod.org forums - LBM: Theory</title>
        <description>Discussions of any kind around the lattice Boltzmann method.</description>
        <link>http://www.lbmethod.org/forum/list.php?3</link>
        <lastBuildDate>Thu, 09 Sep 2010 05:33:02 +0200</lastBuildDate>
        <generator>Phorum 5.2.6a</generator>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2939,2939#msg-2939</guid>
            <title>Bed Elevation? (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2939,2939#msg-2939</link>
            <description><![CDATA[ I am trying to apply LBE to shallow water flow and I want to test it for a case where there is a small bump in a channel for the subcritical flow. Does anyone know how should I introduce the bed elevation in the model? I am using the sample FORTRAN code provided by Zhou at the end of his book.]]></description>
            <dc:creator>mathloverboy</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 06 Sep 2010 08:55:02 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2894,2894#msg-2894</guid>
            <title>Require - Basic LBM Information to start with (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2894,2894#msg-2894</link>
            <description><![CDATA[ Hi ,<br />
<br />
Myself Yogesh , doing Post gratuation (Heat Power) from Pune University . Please give me some information regarding <u class="bbcode">Basics of LBM</u>.]]></description>
            <dc:creator>yogesh</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 26 Aug 2010 10:22:20 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2881,2881#msg-2881</guid>
            <title>about relaxation time and kinematic viscosity (1 reply)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2881,2881#msg-2881</link>
            <description><![CDATA[ Hi everybody,<br />
<br />
I'm now studying single-phase &amp; triple-component LBM and code is almost done, but I've got trouble about value of tau. In my code, tau is calculated by kinematic viscosity, and it is undeniably fact. Result of calculation with species property, values of tau are about 0.8, 1.03, 1.2(H2O, N2, O2, in order). With conditions that I concern, though, calculation reachs just 2195 steps and stop because of negative density value of N2. <br />
<br />
I suddenly hit upon the paper which mentioned about tau that, &quot;value of tau were set to be unity for stability&quot;. I apply it to my code,  now it works 75000 steps and still working, amazingly!!<br />
<br />
So... my question is....<br />
This is CORRECT?? If it is, why? Changing value of tau means that value of kinemativ viscosity become difference and will not equivalent with real property. Can it be suitable...?<br />
<br />
Please, help me :)<br />
<br />
<br />
best regard,<br />
K. N. KIM<br />
<br />
<br />
P.S. <br />
1. Base of my model is paper of Luo and Girimaji[Phys. Rev. E 67, 036302(2003)] and McCracken and Abraham[Phys. Rev. E.71,046704(2005)]<br />
2. I'm really thankful for this forum. Really helpful to me :)]]></description>
            <dc:creator>Kasabian</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 02 Sep 2010 16:24:48 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2864,2864#msg-2864</guid>
            <title>Initialization for DNS - channel case (1 reply)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2864,2864#msg-2864</link>
            <description><![CDATA[ Hi all,<br />
<br />
I have problem in initialization in channel case using LBM. My LBM channel case is working very well for laminar flow but when i try to change to DNS it is not working. I guess my initialization method for DNS is wrong. Do you have any idea how to overcome this problem?<br />
<br />
Please Help<br />
<br />
Regards,<br />
Min]]></description>
            <dc:creator>cn_limin</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 12 Aug 2010 23:52:32 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2855,2855#msg-2855</guid>
            <title>Viscoelastic Boundary Condition for D2Q9 Lattice (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2855,2855#msg-2855</link>
            <description><![CDATA[ Hi,<br />
<br />
I reimplemented the viscoelastic LBM described in <a rel="nofollow"  href="http://library.epfl.ch/en/theses/?nr=4505">Malaspinas: 'LBM for the Simulation of Viscoelastic Fluids'</a>. My aim is to use the D2Q9 lattice for the whole LB domain. Almost all equations can be adapted to this lattice except the equations of the conformation tensor boundary condition described in section 5.2.3.<br />
<br />
Now I am searching for a conformation tensor boundary condition for the D2Q9 lattice.<br />
<br />
Kind regards,<br />
Matthias]]></description>
            <dc:creator>mm_911</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Tue, 10 Aug 2010 15:00:49 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2854,2854#msg-2854</guid>
            <title>scaling in CUDA !! (6 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2854,2854#msg-2854</link>
            <description><![CDATA[ Hi,<br />
<br />
I have a working LBM code written in CUDA for GPU's. The code works fine for nodes (N) less than 7,000. The output is simply Nan When I try to simulate for bigger domain.<br />
<br />
I have observed this behavior both on Linux and Windows based machine. <br />
I understand that each CUDA kernel call should be shorter than the frame rate of GPU, which may be 1/100th of a second.<br />
<br />
Is there a way to get around this limitation of frame rate. How can I simulate for larger domain using GPU?<br />
<br />
Following is kernel call inside the time loop. I have tried both 1-D and 2-D blocks.<br />
<br />
  int blockSize = 256;<br />
  int nBlocks = N/blockSize + (N%blockSize == 0?0:1);<br />
<br />
 for(t=1; t&lt;=frame_rate*num_frame; t++)<br />
  {<br />
<br />
  // calculation on device:<br />
<br />
  collision &lt;&lt;&lt; nBlocks, blockSize &gt;&gt;&gt; (f0_d, rho0_d, ux0_d, uy0_d, is_solid_d, ns_d, tau0, F_gr_d);<br />
<br />
  streaming &lt;&lt;&lt; nBlocks, blockSize &gt;&gt;&gt; (f0_d, ftemp0_d);<br />
  bcs_fluid &lt;&lt;&lt; nBlocks, blockSize &gt;&gt;&gt; (f0_d, rho0_in, rho0_out, ux_in);<br />
  macro_vars &lt;&lt;&lt; nBlocks, blockSize &gt;&gt;&gt; (f0_d, rho0_d, ux0_d, uy0_d);<br />
<br />
}<br />
<br />
Do I need to make asynchronous call for kernels?.<br />
<br />
Thanks<br />
Shadab]]></description>
            <dc:creator>shadab</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 26 Aug 2010 14:36:34 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2829,2829#msg-2829</guid>
            <title>moving boundary and LBM-DEM coupling (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2829,2829#msg-2829</link>
            <description><![CDATA[ Dear everyone, <br />
<br />
     Any people here  are interested in coupling of LBM and DEM ? <br />
<br />
     Any people  tried the boundary condition of  D R Nobleand JR Torrczynski ( int j mod phys, c, 9:1189-1201, 1998 ) ,<br />
<br />
      or  B.K. Cook ( engineering computations, 21:151-168,2004),<br />
<br />
       or Y T Feng ( int j numer method engrg 72,1111-1134,  2007)     ?   <br />
<br />
<br />
    I 'd like to communicate with you.]]></description>
            <dc:creator>ycwang</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 05 Aug 2010 08:36:40 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2827,2827#msg-2827</guid>
            <title>beta factor (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2827,2827#msg-2827</link>
            <description><![CDATA[ I am doing porous media simulation to calculate permeability and beta (Forchheimer) factor. So far my permeabilities are close to values I compare them with. However, the beta factors are off.  I am wondering if anybody has tried this using the LBM and if the beta factors came out well.]]></description>
            <dc:creator>Chuckdee</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 05 Aug 2010 06:53:07 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2824,2824#msg-2824</guid>
            <title>circshift in cavity.m (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2824,2824#msg-2824</link>
            <description><![CDATA[ Dear all,<br />
I'm new in the field of LBM and I started looking at the code cavity.m given in <a rel="nofollow"  href="http://www.lbmethod.org/numerics:matlab_samples">lbmethod</a>.<br />
I have a doubt concerning the streaming step. As far as I understand the matlab command circshift seems to impose not only the streaming but also a &quot;periodicity&quot; between the left-right and and top-bottom sides. As matter of fact the results provided by replacing<br /><pre class="bbcode">
%STREAMING STEP
for i=1:9 
    fIn(i,:,: ) = circshift(fOut(i,:,: ), [0,cx(i),cy(i)]); 
end</pre>
with<br /><pre class="bbcode">
%STREAMING STEP
% implementation of the streaming without &quot;periodicity&quot;
fIn(2, 2:lx, : ) = fOut(2, 1:lx-1,: ); 
fIn(3, :, 2:ly) = fOut(3, :, 1:ly-1); 
fIn(4, 1:lx-1, :) = fOut(4,  2:lx , :); 
fIn(6, 2:lx, 2:ly) = fOut(6,  1:lx-1, 1:ly-1); 
fIn(5, :,  1:ly-1 ) = fOut(5,  :,  2:ly); 
fIn(7, 1:lx-1,  2:ly) = fOut(7, 2:lx , 1:ly-1 );
fIn(8, 1:lx-1, 1:ly-1) = fOut(8,  2:lx , 2:ly); 
fIn(9, 2:lx , 1:ly-1) = fOut(9,    1:lx-1, 2:ly);</pre>
are not equivalent. I don't understand why we need this &quot;periodicity&quot; all over the boundaries if we already have already imposed the corresponding boundary conditions. Is this related to the full bounce back condition? How?<br />
<br />
Thanks a lot]]></description>
            <dc:creator>mast</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Wed, 04 Aug 2010 11:56:19 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2815,2815#msg-2815</guid>
            <title>2nd order convergence of Poiseuille using BGK (3 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2815,2815#msg-2815</link>
            <description><![CDATA[ Dear all,<br />
<br />
I have been trying to reproduce the second order convergence rate with the lattice spacing of a 2D poiseuille flow using BGK, velocity and pressure Zou-He BCs at inlet and outlet and mid-way bounce back at the walls.<br />
<br />
Taking as a starting point the J.Latt's code [<a rel="nofollow"  href="http://www.lbmethod.org/_media/numerics:cylinder.m">www.lbmethod.org</a>], I have fixed tau=1, uMax= 0.1 and computed the norm of the difference of the solution with the analytical solution. The rate of convergence I compute is lower than 1st order.<br />
<br />
If uMax is fixed to a lower value (smaller Ma), the error is lower, but still the convergence rate is lower than 1st order.<br />
<br />
I would like to ask the community if there is anyone that has reproduced the 2nd order convergence of a 2D poiseuille flow using Zou-He BCs at the inlet and outlet and BB conditions at the walls and who is willing to give some tips on where I could have gone wrong.<br />
<br />
Thank you in advance for all comments.<br />
<br />
Kind regards,<br />
Yulian<br />
<br />
Two additional comments:<br />
- a special tratment of the corner nodes could perhaps improve a bit the situation, but in the original article of Zou-He (Phys Fluids, 1997), p.1595, right column explicitly say that they do not specially treat them<br />
- as Alex has pointed out in the [<a rel="nofollow"  href="http://www.lbmethod.org/forum/read.php?3,2124,2150#msg-2150">www.lbmethod.org</a>] post, the effective height is somehow different from the computed one for a midway BB, but even though, the correction is quite small for tau=1 and decreasing rapidly by refining the lattice]]></description>
            <dc:creator>ulian71</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 09 Aug 2010 12:17:18 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2804,2804#msg-2804</guid>
            <title>Lattice units to physical units - permeability tutorial (7 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2804,2804#msg-2804</link>
            <description><![CDATA[ Hi,<br />
<br />
I have gone through the other threads about lattice unit conversion and the document lbunits.pdf still cant figure this out, any help would be appreciated .<br />
<br />
For the permeability tutorial case with nx = 48 ny = 64 nz = 64<br />
<br />
The output in lattice units <br />
<br />
Average velocity (meanU, darcy velocity)   = 0.000177152<br />
Lattice viscosity nu                                  = 0.166667<br />
Grad P                                                    = 1.06383e-06<br />
Permeability                                            = 27.7539<br />
Reynolds Number  meanU*(nx-1)/ nu         = 0.049957<br />
<br />
Can't yet figure out how to get the physical velocity. pressure, viscoity...]]></description>
            <dc:creator>ananjund</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 09 Aug 2010 12:09:42 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2802,2802#msg-2802</guid>
            <title>Macroscopic and microscopic behavior (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2802,2802#msg-2802</link>
            <description><![CDATA[ Hi, LGCA and LBM are considered as microscopic and macroscopic in some aspects, can any one explain how these two behaviors appear in these methods. <br />
<br />
and <br />
<br />
what is the basic difference in LGCA and LBM ??<br />
<br />
Regards]]></description>
            <dc:creator>adnan_jahangir</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Sat, 31 Jul 2010 15:30:44 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2796,2796#msg-2796</guid>
            <title>Thermal LBM and two-phase flow question (3 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2796,2796#msg-2796</link>
            <description><![CDATA[ Hello everybody!<br />
<br />
I am trying to simulate the collision and solidification of liquid iron droplets on a cooled iron surface with a custom coupled Lattice Boltzmann method.<br />
It consists of: <br />
1. An isothermal model to handle the density and velocity fields (The model by <a rel="nofollow"  href="http://linkinghub.elsevier.com/retrieve/pii/S0017931004001528">Inamuro et al.</a> is used) <br />
2. An additional thermal model for handling the temperature evolution (passive scalar method such as in Yuan/Schaefer Part <a rel="nofollow"  href="http://www.engr.pitt.edu/mechanical/faculty-staff/schaefer/papers/YS2006c.pdf">One</a> and <a rel="nofollow"  href="http://www.engr.pitt.edu/mechanical/faculty-staff/schaefer/papers/YS2006b.pdf">Two</a>)<br />
The coupling of the two models is done using the velocity and a body force term as suggested by Yuan/Schaefer.<br />
<br />
The original work of Yuan/Schaefer couples the thermal model with a different isothermal model, but to me it seems like any two-phase model can be used.<br />
<br />
The hot droplets fall through the air onto the surface and are supposed to cool down there very quickly and solidify. From what I understand the droplets should cool down much faster when in contact with the surface than when they are in mid-air. That's because of the different thermal conductivities of air (0.0262 W/ K m) and iron (80 W/ K m). However the thermal diffusivities of air and iron are nearly the same (around 0.000021 m^2/s). And since the thermal diffusivity is the only parameter of the thermal LBM by Yuan/Schaefer it is impossible for me to simulate the behaviour of quicker thermal transport in metal compared to air. My original idea was to modify the relaxation time tauT of the thermal model depending on the density, but that is not consistent with the equation:<br />
<br />
alpha=1/3(tauT-1/2)<br />
<br />
where alpha is the thermal diffusivity. This equation can be found in various papers about thermal lattice boltzmann methods.<br />
<br />
Putting it all together, my questions are as follows:<br />
1. Does anybody have an idea why the above equation holds?<br />
2. Does anybody have an idea how to incorporate different thermal conductivities for the two-phases without modifying tauT?<br />
3. Has anybody done any successful simulations of two-phase thermal flow?<br />
<br />
I know I didn't go into much detail yet, but maybe you can already answer some questions. If not, I can go more specific ;)<br />
<br />
Any help is greatly appreciated!<br />
<br />
Best regards,<br />
Tobi]]></description>
            <dc:creator>ConfusedTobi</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 05 Aug 2010 12:20:19 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2791,2791#msg-2791</guid>
            <title>Adsorption (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2791,2791#msg-2791</link>
            <description><![CDATA[ Hello everyone,<br />
I want to simulate adsorption of Nitrogen from air flow in a packed bed of Zeolit (Pressure Swing Adsorption of Nitrogen) with Lattice Boltzmann Method. <br />
Can anyone help me for this problem?<br />
How to one can model adsorption process in LBM?<br />
 Best Regards]]></description>
            <dc:creator>mehdi</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 26 Jul 2010 15:36:45 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2788,2788#msg-2788</guid>
            <title>Body forces in MRT (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2788,2788#msg-2788</link>
            <description><![CDATA[ Hello, LBMers,<br />
<br />
I'm working on force implement in MRT. I tested Luo and Lallemand's (PRE, 2003) scheme. However, the maximum velocity is not correct in Poiseuille flow. I'm not sure where is wrong...<br />
<br />
What I did is as follows,<br />
<br />
Collision:<br />
<br /><pre class="bbcode">
relaxation paramters:
  s0=s3=s5=0.0;   s1=1.64;   s2=1.54;   s4=s6=1.7;   s7=s8=1.0/tau;

macroscopic value:
  rho=\sum f_i;
  <strong class="bbcode">j</strong> = \sum f_i*e_i + 0.5 * <strong class="bbcode">F</strong>;  // flux, updated with half force 

equilibrium values of the moments:
  m_eq0= rho; m_eq1= -2.0*rho+3.0*(j_x^2+j_y^2)/rho; ....

transfer from population space to moment space:
  <strong class="bbcode">m</strong> =  <strong class="bbcode">T f</strong>;  // T is the matrix

collsion in moment space:
  
 <strong class="bbcode">m*</strong>= s (<strong class="bbcode">m_eq</strong>-<strong class="bbcode">m</strong>);

j&quot; = j'+0.5*<strong class="bbcode">F</strong>; // another half part of the force.

inverse transfer from moment space to population space:
  <strong class="bbcode">f *</strong>=<strong class="bbcode">T^{-1} m*</strong>

End of the collision step</pre>
<br />
Thank you very much.]]></description>
            <dc:creator>phoenixchen903</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 26 Jul 2010 09:10:46 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2783,2783#msg-2783</guid>
            <title>Oxford LB workshop (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2783,2783#msg-2783</link>
            <description><![CDATA[ Dear All,<br />
<br />
Please could I draw your attention to the Oxford Lattice Boltzmann Workshop which will be held between 13th-14th September in New College, Oxford. The event will begin with a tutorial session given by Dr. Paul Dellar and will continue with a series of lectures by members of the LB community. A website for the workshop, including an on-line registration form, can be found here:<br />
<br />
[<a rel="nofollow"  href="http://www.maths.ox.ac.uk/groups/occam/forthcoming-events/occam-lattice-boltzmann-workshop">www.maths.ox.ac.uk</a>]<br />
<br />
If you have any further questions then please get in touch (reis@maths.ox.ac.uk). Student support is available and I encourage those who are interested to get get in touch with me before registering.<br />
<br />
I look forward to welcoming you to Oxford,<br />
Best wishes,<br />
Tim]]></description>
            <dc:creator>pleb01</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Wed, 25 Aug 2010 12:12:56 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2773,2773#msg-2773</guid>
            <title>anisotropic viscosity (8 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2773,2773#msg-2773</link>
            <description><![CDATA[ Hi,<br />
<br />
is there any way how to incorporate anisotropic viscosity into LBM? I would like to simulate liquid with fibers also on macroscopic scale (without modeling the fibers) which makes the fluid anisotropic (different viscosity in different directions).<br />
<br />
Best regards,<br />
Oldrich Svec]]></description>
            <dc:creator>czb</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 05 Aug 2010 07:30:34 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2757,2757#msg-2757</guid>
            <title>HOW TO INCORPORATE ENTRY LENGTH EFFECTS IN POISEUILLE FLOW (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2757,2757#msg-2757</link>
            <description><![CDATA[ HI,<br />
     FRIENDS  i am just beginner of LBM.I have a doubt that how to specify boundary condition for the problem like  &quot;with some inlet velocity fluid is entering into the two parallel plates, both are fixed,driven by a pressure gradient&quot;<br />
<br />
<br />
                                                                  Thank you  Friends]]></description>
            <dc:creator>haribabu pasam</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Fri, 16 Jul 2010 08:26:24 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2751,2751#msg-2751</guid>
            <title>routine works only for certain numbers of processors (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2751,2751#msg-2751</link>
            <description><![CDATA[ Hallo,<br />
<br />
I am looking for the steady state of a two-dimensional flow. The routine I have written computes the maximal relative change of the velocity profile in a specified domain. Everything works fine as long as I run the program on 1, 2 or 4 processors. If I use a different number, something breaks down and I get meaningless results. The code is below, any ideas? Thanks a lot.<br />
<br />
Simon<br />
<br /><pre class="bbcode">
bool steadyState2Dc ( MultiBlockLattice2D&lt;T,DESCRIPTOR&gt;&amp; lattice, Box2D domain, T threshold, plint component, T&amp; diff) {
    static MultiTensorField2D&lt;T,2&gt; uOld(domain.x1+1, domain.y1+1);
    MultiTensorField2D&lt;T,2&gt; uNew(domain.x1+1, domain.y1+1);
    computeVelocity(lattice, uNew, domain);
    subtractInPlace(uOld, uNew, domain);
    divideInPlace(uOld, uNew, domain);

    T diffmax = fabs(computeMax(*extractComponent(uOld, domain, component)));
    T diffmin = fabs(computeMin(*extractComponent(uOld, domain, component)));

    uOld = uNew;
    
    diff = (diffmax &lt;= diffmin) ? diffmin : diffmax;

    return (diff &lt; threshold);
}</pre>]]></description>
            <dc:creator>simon</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 15 Jul 2010 12:31:13 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2745,2745#msg-2745</guid>
            <title>Choice of discretization to ensure stability (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2745,2745#msg-2745</link>
            <description><![CDATA[ Hello,<br />
<br />
I am performing 3d simulations in a pipe with a contraction which causes a nine-fold increase in velocity. The Reynolds number is 500 based on the velocity and diameter inside the contraction. Due to the shape of the domain the diameter of the contraction is about 1 hundredth of the length of the domain. <br />
<br />
My problem is that the lattice viscosity (Nu_LB) is determined uniquely by the diameter (d_LB) and mean velocity (u_LB) inside the contraction by <br />
<br />
Nu_LB = u_LB*d_LB/Re.<br />
<br />
As u_LB can't be above 0.1 and even on a very large grid d_LB is only 20 lattice units this leads to very low values of Nu and therefore a relaxation parameter very close to 0.5 (max 0.512). This was OK using a D3Q19 lattice and I got some reasonable results but the simulations fairly quickly become unstable on a D3Q15 lattice and almost immediately on a D3Q27 lattice. I'd be really grateful if anyone has any ideas about how I might be get a stable simulation without using an unreasonably large grid.<br />
<br />
Thanks,<br />
Alex White<br />
<br />
p.s. I want to use the D3Q27 and D3Q15 lattices to see if they will correct some problems which I saw in the results from the D3Q19 lattice.]]></description>
            <dc:creator>Alex_W</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Wed, 18 Aug 2010 13:57:08 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2728,2728#msg-2728</guid>
            <title>wall velocity in Halfway bonce back (1 reply)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2728,2728#msg-2728</link>
            <description><![CDATA[ Hi Dear friends<br />
I have solved a simple 2D channel flow and used halfway bounce back for the walls.<br />
I used uniform grid with dx=dy=1(lu) for all nodes. <br />
you can see shcematic for showing bounceback:(vertical nodes on the upper wall are shown by numbers)<br />
  <br />
    1<br />
- -|- -physical boundary(upper wall)<br />
    2<br />
    |<br />
    3<br />
    :<br />
    :   <br />
<br />
by halfway bounce back:(upperwall: y=1): f(x,1,i)=f(x,1,opp(i))<br />
and perform other steps of LB algorithm.<br />
At the end of program we have (for example at x=somewhere in fully developed region): u(x,1)=-0.16 and u(x,2)=0.05 ,u(x,3)=0.15<br />
But we know that the node number 1 =(x,1) is located at distance of dx/2 from the wall and u(x,1) is not real and its velocity<br />
useless. But for presentation the velocity profile the first node we need is the wall node that is located<br />
between node 1 &amp; 2 in the figure and this node doesn't contribute in our calculation.<br />
So how we calculate the wall velocity(that should be very close to 0). Can we extrapolate it<br />
by using u(x,2) &amp; u(x,3)?<br />
or we should consider the wall node in a grid with dx/2 distance between first and second node?<br />
 <br />
Is the my procedure right?]]></description>
            <dc:creator>jav1984</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Wed, 14 Jul 2010 16:25:25 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2725,2725#msg-2725</guid>
            <title>Acoustic waves (3 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2725,2725#msg-2725</link>
            <description><![CDATA[ Hello,<br />
<br />
I'd like to use LBM to model the propagation of acoustic waves in a microfluidic channel. In my case study, I have a kind of resonance chamber of about 2x2 mm and 500 um in height, and a bunch of piezoactuated tips that push against the bottom of the chamber. My question is, is a simple BGK scheme adapted to model this kind of effects? More specifically, what would be the Reynolds number in that case, and what would it depend on?<br />
<br />
Thanks,<br />
<br />
Greg]]></description>
            <dc:creator>gmermoud</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Tue, 20 Jul 2010 09:57:28 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2720,2720#msg-2720</guid>
            <title>Pressure BC for chamber (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2720,2720#msg-2720</link>
            <description><![CDATA[ Dear All,<br />
<br />
I wanted to simulated flow in a chamber with one inlet and outlet at the top surface. My domain is D2Q9.<br />
I am using forcing term to include pressure (as mentioned in the forum)<br />
my question is how to set pressure at inlet and though out the domain?<br />
does it be like gradP = 0.5*rho*U^2?<br />
and for inside domain gradP= 0.5*rho*U^2/(distance between inlet &amp; outlet)?<br />
<br />
Please help me<br />
Thanks in Advance]]></description>
            <dc:creator>winter</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Tue, 06 Jul 2010 07:10:04 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2719,2719#msg-2719</guid>
            <title>Ello! (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2719,2719#msg-2719</link>
            <description><![CDATA[ I've just started out to learn about idea from thread. You think that human being can learn as fast some super computer? Share your opinion! :) I've found out that i got highly motivated by reading this stuff! Looking forward your opinions... hope that &quot;Lattice Boltzmann&quot; is good place for my first topic!]]></description>
            <dc:creator>majkiman</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 05 Jul 2010 23:51:04 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2701,2701#msg-2701</guid>
            <title>interaction potential  with shan-chen mode (4 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2701,2701#msg-2701</link>
            <description><![CDATA[ Dear friends:<br />
       we need to caculate the interaction fore as we use the shan-chen mode to simulate the single component multiphase flows. When x+e is outside the fluid domain,then how to determain the interaction potential ?  I need your kind help and best regards<br />
 <br />
Yours <br />
<br />
Gongyou Liang]]></description>
            <dc:creator>lianggy</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Tue, 03 Aug 2010 20:10:32 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2693,2693#msg-2693</guid>
            <title>2D Channel: Poiseuille Flow (1 reply)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2693,2693#msg-2693</link>
            <description><![CDATA[ Hi,<br />
<br />
This post is closely related to: [<a rel="nofollow"  href="http://www.lbmethod.org/forum/read.php?3,2479,2508#msg-2508">www.lbmethod.org</a>]<br />
<br />
I'm using the following setup:<br />
<br />
i = inlet <br />
o = outlet <br />
X = WALL <br />
&lt;code&gt; <br />
XXXXXXXXXXXXXXXXXXXXXXXXXX<br />
i ________________________o <br />
i ________________________o <br />
i ________________________o <br />
i ________________________o <br />
XXXXXXXXXXXXXXXXXXXXXXXXXX<br />
<br />
I put a pressure boundary at inlet and outlet, the Zou/ He condition, as described in the above link.<br />
<br />
I'm seeking a clarification, when I implement this type of method, I still have an increase in my x velocity at the outlet compared to the inlet, and some y direction velocity near the outlet both up and down.<br />
<br />
Should the velocity not be uniform from inlet to outlet?<br />
<br />
Does anyone else have this problem?<br />
<br />
Thanks in advance for any advice.]]></description>
            <dc:creator>westcoast</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Mon, 28 Jun 2010 21:07:04 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2685,2685#msg-2685</guid>
            <title>Lattice Units Calculations, problem with density (10 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2685,2685#msg-2685</link>
            <description><![CDATA[ Hello,<br />
I am a new user of Lattice Boltzmann experiencing the same problems, I guess, that you had. I read all the thread about lattice unit calculations, unfortunately, I am still not convinced about what I am doing, hence I would like to have a double check with you, that already solved this problem. I read the guide made by Jonas Latt, but I still have doubts. Here I report my example, if you can, please read, it and tell me if it is correct. <br />
The characteristic diameter is indicated with D, with u the velocity, g the gravity acceleration, rho the density and nu the viscosity, as usual _LB in lattice Boltzmann units and _P in physical ones.<br />
<br />
Reynolds=100, D_LB=70 [lu], D_P=300E-6 [m], u_LB=0.025 [lu/ts], u_P=0.35 [m/s], g_P=9.81[m/s^2], rho_P=1000 [kg/m^3], nu_P=1.06E-6[m^2/s].<br />
<br />
dx=D_P/D_LB=300E-6/70=4.3E-6[m/lu]<br />
dt=(u_LB/u_P)*dx=(0.025/0.35)*4.3E-6=3.06E-7[s/ts]<br />
nu_LB=u_LB*D_LB/Re=0.025*70/100=1.75E-2 [lu^2/ts]<br />
tau=(6*nu_LB+1)/2=(6*1.75E-2+1)/2=0.553<br />
g_LB=g_P*dt^2/dx=9.81*3.06E-7^2/4.3E-6=2.15E-7[lu/ts^2]<br />
rho_LB=rho_P*dx^3/1=1000*(4.3E-6)^3=<span style="color: #FF0000">7.90E-14</span>[1/lu^3]<br />
<br />
The last point doesn’t convince me. I am simulating an incompressible flow across a pipe, and I read that you suggest to use rho_LB=1 but it doesn’t fit with this calculations, please, can you tell me where is the error and if the procedure is correct, please?<br />
<br />
Kind Regards Antonio]]></description>
            <dc:creator>mortain</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Tue, 13 Jul 2010 11:50:05 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2680,2680#msg-2680</guid>
            <title>Chapman-Enskog (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2680,2680#msg-2680</link>
            <description><![CDATA[ Hi,<br />
I am rather new to the LB method and besides reading and trying out some things about it,<br />
I also need to catch up on some of the mathematics involved. Could someone point me<br />
to a place where I can learn about the Chapman-Enskog expansion? What sort of book may contain that, would that be functional analysis?<br />
<br />
Thanks for any hint,<br />
Christian]]></description>
            <dc:creator>cg</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Fri, 25 Jun 2010 11:53:37 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2676,2676#msg-2676</guid>
            <title>Flow Past circular cylinder no vortex (2 replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2676,2676#msg-2676</link>
            <description><![CDATA[ Hi folks,<br />
      I am using the same code posted on the site, (flow past circular cylinder). I have simulated the result for Re=20,40 but I am getting no vortex formation in the rear part of the cylinder, I dunno why...<br />
xDim = 600<br />
yDim= 250<br />
obst X=xDim/5<br />
obstY=yDim/2<br />
obstR=yDim/10+1<br />
tMax=10,000<br />
Re=50.d0<br />
     Kindly let me know why I am not getting the vortices. The code is D2Q9. omega was 1.7 for this case approximately. The code is the same f90 code posted on the side for flow past cylinder. Kindly let me know error in the input data.<br />
Regards<br />
Prateek]]></description>
            <dc:creator>sharmap</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 24 Jun 2010 16:38:07 +0200</pubDate>
        </item>
        <item>
            <guid>http://www.lbmethod.org/forum/read.php?3,2675,2675#msg-2675</guid>
            <title>Opportunities available at Laurentian University ! (no replies)</title>
            <link>http://www.lbmethod.org/forum/read.php?3,2675,2675#msg-2675</link>
            <description><![CDATA[ Hi LBMers,<br />
 <br />
Graduate (MSc or PhD) or postdoc opportunities are available with Dr. Zhang in School of Engineering at Laurentian University, Canada. Potential topics include biofluidics, microfluidics, microscopic interfacial phenomena, colloidal systems, computational physics, and simulation technologies. Detailes of Dr. Zhang's background and research activities can be found at  <u class="bbcode">HTTP://nanotech.Calgary.ca/~Zhang/Zhang_CV.html</u>. <br />
 <br />
Candidates interested in these areas with backgrounds in engineering and science are encouraged to apply by contacting Dr. Zhang at <u class="bbcode">jzhang@laurentian.ca</u>.<br />
 <br />
Thanks for passing it around and please let me know if anything I can help.<br />
<br />
Best regards!<br />
 <br />
KaiSun]]></description>
            <dc:creator>kaisun</dc:creator>
            <category>LBM: Theory</category>
            <pubDate>Thu, 24 Jun 2010 13:28:46 +0200</pubDate>
        </item>
    </channel>
</rss>
