Coordinates of GFS file

[Hiroumi Tani]

Hello Stephane,

The x-y-z coordinate of gfs files looks different from that in the programs.
I ran "atomisation.c" and output the gsf files to see by gfsview.
In the program, the inflow boundary of the pulsed jet is specified to the x_min plane. 
However, the jet came from the z_max plane when I visualized it  by gfsview. Why do they mismatch?

One more thing. Gerris has an ability to generate the tecplot format data.
Do you have plan to implement  the same ability to Basilisk?

Thanks,
Hiroumi

[Stephane Popinet]

> The x-y-z coordinate of gfs files looks different from that in the programs.
> I ran "atomisation.c" and output the gsf files to see by gfsview.
> In the program, the inflow boundary of the pulsed jet is specified to
> the x_min plane.
> However, the jet came from the z_max plane when I visualized it by
> gfsview. Why do they mismatch?

Please see:

https://groups.google.com/forum/#!topic/basilisk-fr/yGedVYOCz8E
https://groups.google.com/forum/#!topic/basilisk-fr/A6TKu16wbsQ

> One more thing. Gerris has an ability to generate the tecplot format data.
> Do you have plan to implement the same ability to Basilisk?

Not personally, because I don't have/use tecplot, if someone wants to do
it I would of course be happy to include the code.

cheers

Stephane

[Hiroumi Tani]

Hello Stephane,

OK. I understood the difference in the ordering.

>Not personally, because I don't have/use tecplot, if someone wants to do
>it I would of course be happy to include the code.

I and my co-workers definitely want to use tecplot for basilisk.  We wish you include it!

Thanks,
Hiroumi

2016年8月25日木曜日 7時39分53秒 UTC+9 Stephane Popinet:

[Vitor Vilela]

Dear Hiroumi,

I understood the sentence "if someone wants to do
it I would of course be happy to include the code" a little bit different from you. I think it means that Stephane will (just) add the code to an official Basilisk release after any other person works on it.

Regards,
Vitor Vilela.

[Hiroumi Tani]

Hello Vitor,

Oh, I misunderstood it. I am trying to program the code for tecplot. Hope I can send it to all.

Thanks,
Hiroumi

2016年9月1日木曜日 0時29分33秒 UTC+9 Vitor Vilela:

[22960...@qq.com]

Dear Hiroumi Tani

  I am also trying to write an Output_Tecplot code, but i have trouble in writing the connectivity list. 

  It's trivial to write an Tecplot output on Cartesian grid, just add :

  VARIABLES = "X","Y","C" 

  ZONE T = "Floor", I = num_of_point_along_x , J = num_of_point_along_y, F = POINT

  at the beginning of the output file, then we can open it in Tecplot.

  However, when we want to write an output on adaptive grid, we should know the number of nodes, as well as number of Elements. And the most important thing is to write the connectivity list to show how these nodes are connected to form elements. The format should be :

   --------------------------------------

   VARIABLES = "X","Y","C"

   ZONE N = num_of_nodes, E = num_of_element, F = FEPOINT, ET = QUADRILATERAL

   --------------------------------------

   Our Data
   --------------------------------------

   Connectivity List

   --------------------------------------

   

    I don't know how to write the connectivity list. Have you solved this problem yet? I am looking forward to your reply.

    Thanks,

    Ray 

  

在 2016年9月7日星期三 UTC+8下午9:12:22，Hiroumi Tani写道：

[Hiroumi Tani]

Hello Ray,

A test code and header file were uploaded to the sandbox (http://basilisk.fr/sandbox/hiroumi/tecplot/) to show my way to output data in tecplot format. To be honest, I hesitated to open them because they are not sophisticated yet. If I understand correctly, Tecplot cannot visualize variables defined at the cell faces. So, my code shows how to transfer the data at cell center and vertex to the Tecplot format. I'm developing codes with Cantera library to simulate reactive flows. So, you will see parts of "#if CANTERA ~ #endif" in the tecplot.h, but you can ignore them.

Best,
Hiroumi

[Hossain Chizari]

Hello all,

The code in the (http://basilisk.fr/sandbox/hiroumi/tecplot/) is using the "tecio" library if I am not mistaken. For a simpler data output (ASCII) in Tecplot, I just put everything in the center of the cell. Then, I use a simple header file,

variables = x, y, f, u, v, ...

ZONE

N = [point numbers], E = [cell numbers]

DATAPACKING = BLOCK, ZONETYPE = FEQUADRILATERAL

VARLOCATION = (NODAL, NODAL, CELLCENTERED, CELLCENTERED, ...) \\ only x and y are NODAL, others are CELLCENTERED

SOLUTIONTIME = [time of the solution]

The [point numbers] will be the [cell numbers] multiply by 4 (N=4*E). In this sense, you will have a cell connectivity itself and do not need to engage the complexity of VTK format. For each cell there will be 4 points; thus, if it is doing in order the connectivity will be

1 2 3 4

5 6 7 8

...

Note that I use BLOCK format here, so we shall have separate loops for each variable.

And, lastly, for the "vof" interface finding I did these steps:

1. reconstruction(f, n, alpha); [http://basilisk.fr/src/geometry.h]

2. Find the interface points based on [n] and [alpha]. At the same time, find the number of intersection edges.

3. The header will be

variables = x, y

ZONE

N = 3*[number of intersection edges], E = [number of intersection edges], DATAPACKING = BLOCK, ZONETYPE = FETRIANGLE

PASSIVEVARLIST = [3-10]

SOLUTIONTIME = [time of the solution]

4. Then, for each element, the first point, then the second point and then the first point again should be written to close the element for Tecplot.

5. The connectivity is the easy part,

1 2 3

4 5 6

...

Best to all

Regards

Hossain