CheckKeyword: Unlisted keyword:

[jjmceld]

Hello to all of you,

I have been trying to put to work the advection diffusion solver for a minimum working example of just diffusion transport, this is with NO convection, in a cube with the respective diffusive constant and body force calculated to have the apropiated diffusive length. However, the solver log file shows two Warnings which are,

ResultOutputSolver: Saving in unstructured VTK XML (.vtu) format
WARNING:: WriteVtuFile: Are there really no scalars or vectors?
ResultOutputSolver: -------------------------------------
SaveScalars: -----------------------------------------
SaveScalars: Saving scalar values of various kinds
SaveScalars: No operator given for variable: phole
WARNING:: SaveScalars: Found no values to save


There are just two boundaries with Dirichlet conditions, The top and the bottom ones. The remaining boundaries only have fluxes equal to zero. There is NO dat file (phole.dat) of the expected values of the depended variable of the diffusion equation.

As it can be seen in the Paraview image the vtu file just have two values, zero and the one of the top boundarie.

¿What could be the mistakes that do not allow to calculate the concentration values through all the cube?

Thank you and looking forward to an answer.

John

Difussion_cube.msh

(9.69 KiB) Downloaded 142 times

Difussion_cube.geo

(2.89 KiB) Downloaded 137 times

diffusion_cube_oct_18_2021_ss.sif

(3.14 KiB) Downloaded 148 times

[Rich_B]

diffusion_elmergui.zip

(132.47 KiB) Downloaded 137 times

Hello,

In your sif file in the Simulation section, you define "Post File = case.vtu", and then later add ResultOutputSolver. One should add one or the other commands, but not both. The first command will create a vtu file with contents, the second command will be empty, as you described.

The attachment is an attempt to recreate your example, using ElmerGUI to provide formatting of the various commands.

Please let us know if this helps.

Thanks, Rich.

[raback]

Hi,

Rich is right. The reason is that 95% of time people want to save results as VTU files so it made sense to create a automated logic for the default settings. There are tons of keywords to control the output in more detail which are easier dealt with in a separate solver section. But if you use both strategies at the same time there seems to be some issue related to static variables.

-Peter

[Rich_B]

Hello,

Would it be possible to add a warning if both options are active in the same sif file?

This has come up before several times, and a polite warning to inform the user would be very helpful.

Thanks, Rich.

[raback]

Hi Rich,

Just committed this polite warning.

-Peter

[Rich_B]

Hello Peter,

Great, it works perfectly!

Thank you for adding that warning and preventing addition of two result solvers!

Rich.

[jjmceld]

Hello Rich,

Thank you for your answer. After deleting the ResultOuputSolver and just leaving the command "Post File = case.vtu" en the simulation section I am getting the following result in Paraview.

Diffusion_image_oct_23_2021.png (76.31 KiB) Viewed 2615 times

One main goal is to obtain the dat file either with the "SaveData" "SaveScalars" Procedure or with the "SaveData" "SaveLine" one. With the first one the Solver log gives the same Warning,

WARNING:: SaveScalars: Found no values to save

And it does not save the dat file. With the second procedure, "SaveData" "SaveLine", the result is the following,

Variables in file: Phole_1.dat
Elmer version: 9.0
Elmer compilation date: 2021-07-21
Solver input file: diffusion_cube_oct_23_2021_ss.sif
File started at: 2021/10/23 17:13:38
Number of data nodes for each step
bc nodes: 0
polyline nodes: 0
circle nodes: 0
isocurve nodes: 0
Data on different columns
1: Iteration step
2: Boundary condition
3: Node index
4: coordinate 1
5: coordinate 2
6: coordinate 3
7: phole

All nodes are zero.

¿How can be saved the dat files either with the "SaveData" "SaveScalars" or "SaveData" "SaveLine" procedures?, ¿What is wrong in the code to achieve this purpose?

Thank you and looking forward to an answer,

John

[Rich_B]

Hello,

Good to see that now the solver runs to completion.

Please post your latest set of files, so we can see what is going on.

Thanks, Rich.

[Rich_B]

Hello,

Looking at your diffusion_cube_oct_18_2021_ss.sif file, the body definition section looks like this:

Body 1
Target Bodies(1) = 1
Name = "top_region"
Material = 1
End

Body 2
Target Bodies(1) = 2
Name = "medium_region"
Equation = 1
Material = 1
Body Force = 1
End
......

Try adding Equation = 1 to both top and bottom regions (in other words, all three bodies). The top and bottom boundary conditions won't do much, if top and bottom aren't included in the solution.

To keep the solution from going to infinity, I had to change the body force from "tx/0.1e-8" to "tx/1.0e4".

Attached is another attempt at your model, this time including savedata and saveline.

Please let us know what you find.

Thanks, Rich.

[jjmceld]

Hello Rich,

The sif file I have attached is based on the last sif file that you modified. The goal I am trying to achieve is to obtain just the diffusion transport in body 2. In order to do so, I added a second material, material 2, for body 1 and 3, with a very low Concentration Diffusivity and high density. However, it does not work with these values, it works better with the opposites ones, as shown in the following figure from the vtu file and the sif file.

Difusion_oct_25_2021.png (74.03 KiB) Viewed 2522 times

As it can be seen, the boundaries values of the top and bottum surfaces for body 2 are the right ones. However, the top surface boundary value for body 1 does not have the 0.0 value given and this body seems to have the same concetration throughout all its volume according to the figure.

¿How could be possible to obtain just the difussion transport in body 2 (middle body)?

Thank you and looking forward to an answer,

John