Learning Forum

[Germanchoacr]

Hi I have a Fluent bus exhaust transient simulation with these high residuals:

image.png

[Nikhil N]

Hello
Ansys staff are not permitted to download attachments. However, you can include screenshots in your post itself.
Can you provide more details on the computational domain, mesh, models, and the solver settings?
As you have tried multiple things, can you try refining the mesh and check if that works?

Nikhil

[Germanchoacr]

Hello nnarale thanks for answering. The residuals:
The domain is made with a mesh of about 2.5 million elements, the physics corresponds to a turbulent, transient, multicomponent, incompressible system. What you say about solver I dont know if are solution methods I have in Solution: Methods: scheme: simple...

[Nikhil N]

Hello,

It looks like the residuals are still moving down. Why don't you try extending your simulation for some more iterations and let the residuals stabilize a bit.
Can you also try monitoring the parameter of your interest (for example, pressure or velocity) and not only the residuals and see if that converges (Reaches to a constant value after some iterations)?

Nikhil

[Rob]

Given it's transient you also need to run the model long enough to cover whatever time you're interested in. Why is the model transient and not steady?

[Germanchoacr]

Hello Nikhil and Rob, I extend the simulation more iterations and these are the residuals:
In some point the residuals stabilize a bit until iteration 1100. The continuity residual is still big, is it?

[Nikhil N]

Hello,

It is difficult to comment on the convergence of the solution just on the basis of the residual plots. I would recommended you to plot and monitor the output parameters of you interest. Check if those parameters are reasonable and perhaps are constant (provided that those parameters are not inherently un-steady).

Nikhil

[Germanchoacr]

Nikhil even before I tried monitoring a parameter of interest, the concentration, but it does not appear as a variable in XY plotes:

[Germanchoacr]

[Germanchoacr]

[Nikhil N]

Hello,

If you are using Specie Transport model, 'Species' function should appear in the function's drop down menu. Can you check if the model is turned on because from the latest residual plot that you shared, it looks like you are not solving for the species.
If not X-Y plot, you can define a surface report from 'Report definition' for monitoring species concentration, provided that the Specie Transport model is turned on, with respect to iterations/time steps and plot and monitor it along with the solution.

Nikhil

[phanicmsf]

Thank you.

[Germanchoacr]

Hi Nikhil, your previous hint about refining the mesh, I found this option:

which of available showed criterium should I to use. Other person suggest me reduce cell size, in ansys fluent guide appears a section called Modify Cells dialog box but without instructions

[Germanchoacr]

As refinement criterion I created a cell register for all the domain and after click in calculate appears these changes:
Grid size [original |adapted |change ]
cells [2,494,995 |9,979,980 |+7,484,985 ]
faces [5,105,117 |30,285,321 | +25,180,204 ]

[Nikhil N]

Here you are trying to activate automatic mesh refinement, which helps fluent to refine/coarsen the mesh during a calculation.
What I suggested is to try refining the mesh before you import it in Fluent solver. In simple words, try reducing the size of the cells in your mesh. Note that, this will increase the overall computational time.

Nikhil

[Nikhil N]

Any thoughts on this?

[Germanchoacr]

Hi the simulation reports: solution converged since time step 8 of 2000 programmed, reaching a continuity residual of 1e-5 :
now I have a trouble with a mixture properties
having same properties values than a different previus mixture
and that volume_volume appears with air not with the mixture
Finally appears erased the nitrogen-oxide-air mixture and I don┬┤t find how to create it again...

[Germanchoacr]

now volume_volume appears with the mixture that I edit from other case (the original) but in Inlet exhaust I can not add the NO specie:

[Nikhil N]

If there are 3 species, you just have to define mass fractions for the two. Solver will calculate the mass fraction for the 3rd specie (1 - addition of the two mass fractions).

[Germanchoacr]

Thansk Nikhil I already know that before, but it does not answer my questions
Thanks in advance

[Germanchoacr]

Now in Inlet exhaust I can add the NO specie. I still have a dude with a mixture nitrogen-oxide-air properties having same properties values than a different previus original case mixture and than pure NO...

[Germanchoacr]

I can add de NO specie changing the order of the species of the mixture clicking Edit in name and then the Last Species buttom

[Germanchoacr]

A plane surface of parcial results; this result has not changed from last autosave timesteps.
It suposed to reach this results for other gas, CO:
why results does not change with timestep advance and NO still not get in into the bus?

[Rob]

What concentration of CO relative to NOx have you got? Remember you have convection and diffusion to consider.

[Germanchoacr]

The NO simulation finished with a little change, and without gas inside the bus, and with particulate matter 2.5 and CO yes other persons got the exhaust pollutant inside that, what could happened?

[Rob]

/forum/discussion/38977/how-to-read-interna-external-bus-pollutant-concentration-in-ansys-fluent#latest Duplicate