Here is a summary of the updates as of 5/25/08:

In the files ~/slfcfd-1.1/cfd_gr/fdpost.c and ~/slfcfd-1.1/cfd_gr/fdgetnew.c, I changed the lines from:

                if( incompressible_flag )
                {
                        check = qikreader( matl, name, node, o3, path, streak);
                        if(!check) printf( " Problems with qikreader\n");
                        if(path.new_flag)
                        {
                            for( j = 0; j < path.num_pt0; ++j )
                            {
                                path.pt[j].x = path.pt0[j].x;
                                path.pt[j].y = path.pt0[j].y;
                            }
                        }
                        if(streak.new_flag)
                        {
                            for( j = 0; j < streak.num_pt0; ++j )
                            {
                                streak.pt[j].x = streak.pt0[j].x;
                                streak.pt[j].y = streak.pt0[j].y;
                            }
                        }
                }

to:

                if( incompressible_flag )
                {
                        check = qikreader( matl, name, node, o3, path, streak);
                        if(!check) printf( " Problems with qikreader\n");
                }

                if(path.read_flag)
                {
                        if(path.new_flag)
                        {
                            for( j = 0; j < path.num_pt0; ++j )
                            {
                                path.pt[j].x = path.pt0[j].x;
                                path.pt[j].y = path.pt0[j].y;
                            }
                        }
                }
                if(streak.read_flag)
                {
                        if(streak.new_flag)
                        {
                            for( j = 0; j < streak.num_pt0; ++j )
                            {
                                streak.pt[j].x = streak.pt0[j].x;
                                streak.pt[j].y = streak.pt0[j].y;
                            }
                        }
                }

Look near the lines 853 and 914 in "fdpost.c" and 668 and 729 in "fdgetnew.c. These lines are inside the statements "if( post_flag )" and "if( input_flag )". Note that path and streak are changed to path0 and streak0 for the code within the "if( input_flag )" statement. This change is necessary because when you read in a new input file, the above code will check if there exits corresponding pathline and streakline data.

Here is a summary of the updates as of 4/25/08:

For the Makefile in the main directory ~/slfcfd-1.1/, I modified the make's pertaining to the AMR files, specifically:

clean:
        cd common ;  make clean
        cd compressible ; make clean
        cd AMR ; make
        cd incompressible ; make clean
        cd common_gr ;  make clean
        cd cfd_gr ;  make clean

needs to change to:

clean:
        cd common ;  make clean
        cd compressible ; make clean
        cd AMR ; make clean
        cd incompressible ; make clean
        cd common_gr ;  make clean
        cd cfd_gr ;  make clean

and

clean_science:
        cd common ;  make clean
        cd compressible ; make clean
        cd AMR ; make
        cd incompressible ; make clean

needs to change to:

clean_science:
        cd common ;  make clean
        cd compressible ; make clean
        cd AMR ; make clean
        cd incompressible ; make clean

As you can see, I need to add clean to the make for the AMR code. This shouldn't affect anyone significantly.

Here is a summary of the updates as of 4/16/08:

Version 1.1 is finally here.

You can get it at the download page. The major addition is you can now solve 2-D compressible flows. I had originally planned for this to be a 1.0 release after the 0.75 release prior to this, but the changes are significant enough that I'm jumping to 1.1. Here is a list of the changes:

Science Code Changes

As mentioned before, I have added an incompressible code which uses the 2-D quickest algorithm given by Davis and Moore and which I have modified to keep true upwinding. In addition to being more scientifically true, I think it dramatically improves the results.
Added new boundary conditions for the incompressible code. The most significant new boundary, noslip for obstructions in the flow, has also been added to the compressible code.
Added the ability to calculate streaklines and pathlines. This is true for both the compressible and incompressible code.
In the compressible code, I had to change the way the steps are counted so that now it stops one step less than the 0.75(or 1.0) release. This was done after careful consideration of the way "dt" was added to the time. Slight numerical error would sometimes result in an additional step, so I changed the code such that this error will not affect the results. This changes the previous results though, in that they are often one step less than the calculations of the previous release.

Graphics Code Changes

The graphics code, fdpost, can handle both compressible and incompressible data. And it will display the pathlines and streaklines.

Data Changes

The data format for compressible data is almost the same with the only difference being when you use the new noslip condition(I say more about this below). For incompressible data, the header data at the top is very similar. The major differences are that the SOR convergence criteria is given on the line for time and time step parameters. And for material properties, you only have the Reynolds Number.

As an example, the beginning data looks like:


total nodes, nodes in x, nodes in y for the square obstruction problem at 0 degree
4590     102   45
Time step, number of steps, current step, current time, SOR tolerence
  4.00000000e-02    7200       0   0.00000000e+00  1.0e-4
spatial discretization dx, dy
  2.66000000e-01   2.66000000e-01
Reynolds Number
  1.00000000e+03

For nodal data, the format of nodal indices, boundary condition, and list of numerical value of nodal quantities is the same, except that for incompressible fluids, I only have pressure, velocity x and y, and the velocity magnitude. So unlike the compressible data, there is no column for density, temperature, enthalpy, energy, and mach number. For example:


   i    j    bc           pressure     velocity x   velocity y   velocity magnitude
   0    0    prs0         0.0000e+00   1.0000e+00   0.0000e+00   1.0000e+00
   0    1    ifxm        -4.4819e+00   1.0000e+00   0.0000e+00   1.0000e+00
   0    2    ifxm        -5.7100e+00   1.0000e+00   0.0000e+00   1.0000e+00
   0    3    ifxm        -6.0169e+00   1.0000e+00   0.0000e+00   1.0000e+00
   0    4    ifxm        -6.0954e+00   1.0000e+00   0.0000e+00   1.0000e+00
   0    5    ifxm        -6.1224e+00   1.0000e+00   0.0000e+00   1.0000e+00
....
  21   18    none        -6.6126e+00   1.3931e+00  -1.4570e-01   1.4007e+00
  21   19    none        -6.6786e+00   1.0897e+00  -8.1584e-02   1.0928e+00
  21   20    nsym      1 -6.6981e+00   1.9648e-01  -6.2475e-04   1.9648e-01
  21   21    nsxp      1 -6.6940e+00  -6.4406e-04   8.4037e-02   8.4040e-02
  21   22    nsxp      1 -6.7325e+00   0.0000e+00   6.7075e-02   6.7075e-02
  21   23    nsxp      1 -6.7655e+00   0.0000e+00   4.0722e-02   4.0722e-02
  21   24    nsxpyp    1 -6.8127e+00  -2.9244e-04   2.5593e-04   3.8862e-04
  21   25    none        -6.8144e+00   2.9871e-01   7.4387e-02   3.0784e-01
....
 101   39    fr2xp       -6.5077e+00   1.0044e+00   1.9623e-02   1.0046e+00
 101   40    fr2xp       -6.5061e+00   1.0103e+00   1.4121e-02   1.0104e+00
 101   41    fr2xp       -6.5049e+00   1.0130e+00   9.1262e-03   1.0131e+00
 101   42    fr2xp       -6.5042e+00   1.0139e+00   4.9290e-03   1.0139e+00
 101   43    fr2xp       -6.5038e+00   1.0143e+00   1.1551e-03   1.0143e+00
 101   44    fsyp        -6.5038e+00   1.0143e+00   1.1777e-03   1.0143e+00

Everything is clearly labeled so hopefully, there won't be too much confusion. The only thing to note is that for noslip boundaries, which are given by "ns" followed by the direction (e.g. xp, ym, etc.), there is an integer number which specifies the body to which that particular noslip boundary belongs. SLFCFD assumes every noslip boundary belongs to a body and if this number is not given, SLFCFD will treat all noslip boundaries as belonging to one.

There are now 2 input data files for the pathlines and streaklines named (filename).path and (filename).streak which can be either the beginning points of the pathlines and streaklines or streakline and pathline data from a previous calculation. The data format of both files are exactly the same. These files are also written to be read into "gnuplot" which I will elaborate on more below. You do not need these files to run the problem, but they will enhance the results.

A typical beginning input file for streaklines with 21 seed points will look like:


number of streaklines
21
streakline number, initial x, initial y
   0     2.92600000e+00   4.78800000e+00
   1     2.92600000e+00   4.90770000e+00
   2     2.92600000e+00   5.02740000e+00
   3     2.92600000e+00   5.14710000e+00
   4     2.92600000e+00   5.26680000e+00
   5     2.92600000e+00   5.38650000e+00
....
  15     2.92600000e+00   6.58350000e+00
  16     2.92600000e+00   6.70320000e+00
  17     2.92600000e+00   6.82290000e+00
  18     2.92600000e+00   6.94260000e+00
  19     2.92600000e+00   7.06230000e+00
  20     2.92600000e+00   7.18200000e+00

The initial x and y columns represents the initial points at which the 21 streaklines will be released. If this were a pathline file, they would represent the first point of a pathline.

In the case of a continuing calculation, the streakline data would look like:


  4.64240   4.96404   4.59325   5.06934   4.83451 ...  4.85146   7.69341   ...  4.87150   7.88612
  4.59832   4.96735   4.55100   5.07675   4.78422 ...  4.80221   7.68081   ...  4.82419   7.87526
  4.55568   4.97122   4.51051   5.08458   4.45475 ...  4.75348   7.66732   ...  4.77712   7.86362
  4.51436   4.97540   4.47160   5.09136   4.41986 ...  4.70523   7.65316   ...  4.73028   7.85140
  4.47431   4.97890   4.43415   5.09703   4.38617 ...  4.65747   7.63842   ...  4.68370   7.83866
...

  2.96252   4.79574   2.96225   4.91559   2.96197 ...  2.96301   6.95662   ...  2.96372   7.19617
  2.92600   4.78800   2.92600   4.90770   2.92600 ...  2.92600   6.94260   ...  2.92600   7.18200

Note that there is no beginning descriptive text. Also, there are 42 columns of numbers, one for each of the 21 x coordinates and 21 y coordinates.

Go here to see the previous update page listing the modifications leading to Version 1.1. .