[divergence]

I'm facing weirdest of problem here and I thought that my fellow CFD engineers here at CFD Online might be able to help me on my troublesome task. My situation is as follows:

I'm modelling a 32 core HPC case with Fluent 15.0. The case includes a channel with a static mixer. There is only one inlet and outlet and the mixer is substracted off the fluid volume. I used an UDF to define the inlet profile for CO and O2, so species model was used but no reactions are applied. The UDF seems to work just fine as the inlet surface for CO and O2 is ok. However, to my surprise, I noticed that the flow rates (kg/s) of the UDF variables do not add up when monitoring inlet and outlet surfaces, i.e. gas compounds disappear mysteriously in the flow volume. Yes, it does sound absurd, but I haven't found any place where the gas could have been "buried" in the volume. Moreover, the total mass flow rate check tells me that everything is in order.

In addition to species, I've applied the RSM turbulence model. The inlet is defined as mass flow boundary condition. Heat transfer is taken into account with the walls and also the DO model is used for solving the radiation.

The reason I am writing this as a UDF thread is that this problem only occurs with the case when the UDF is applied. I'm puzzled with the current situation as it seems that the UDF does what is needed (defines the inlet gas profiles), but the weird stuff happens in the flow volume where the UDF should not apply anymore.

If anyone could shed some light into this problem, I would naturally be quite grateful.

[pakk]

However, to my surprise, I noticed that the flow rates (kg/s) of the UDF variables do not add up when monitoring inlet and outlet surfaces, i.e. gas compounds disappear mysteriously in the flow volume.

So, inlet + outlet != 0

Moreover, the total mass flow rate check tells me that everything is in order.

So, inlet + outlet + something else in the table = 0

Just look at the total mass flow rate table to see where your disappearing mass goes to...

[Ema40]

Hi, I have a similar problem. I have an inlet with a mass flow inlet as boundary condition. After some time steps, it disappears!!! I can't understand why!? I have already perform simulations, but this problem is new for me. Do you know why?
Thank you

[divergence]

Unfortunately I did not manage to solve this problem. However, I managed to go around the problem by defining my inlet boundary condition without any UDFs with slight additional alterations, which was not an optimal solution but it was a solution that gave me a chance to continue my work in a tight time frame.

So the whole mystery concerning disappearing species with UDFs remains a perplexing mystery to me. I hope you can find some way to correct things in your calculations.

[Ema40]

Thank you.
I hope!

Regards

[meusha]

Hello everyone,


I actually struggled with a similar problem for weeks. Mass disappearing mysteriously. The boundary conditions had been enforced, I found out after checking. So the mass was not escaping through the boundaries at all.


I noticed that the continuity equation struggled to converge. Even though it was not exactly diverging, the residue hovered over an unacceptably large value. So, I investigated further to determine that ther was a small issue in my model which made it impossible for the continuity equation to converge, and quite rightly so! I wrote a short UDF to address the issue and immediately the problem disappeared.


So, there is some inconsistency in your model, for sure, if you are experiencing similar difficulties. Check your model thoroughly to see if everything makes sense physically. Read the documentation on the models being employed and try to check if anything is amiss.


Kind regards


Eusha