Learning Forum

[Weiqiang Liu]

Hi all,

I am modelling methane catalytic combustion with fluent. Both gas and surface mechanism are included. Also effective diffusion model is incorporated with UDF. 

When the channel diameter is 2.6 and 2.1 mm, reasonable results can be obtained. However, when channel diameter decreased to 1.8 mm, divergence happed after several hundreds of iterations.

I tried to decrease URF and increase mesh quality. However nothing works. I can guarantee that all other settings are the same for different channel diameters. I am wondering why divergence will happen when just geometry is changed for example decreasing channel diameter.

Best regards

Weiqiang

[Rob]

If it takes a while to diverge what does the flow field look like before it fails?  Check the mesh quality (ie is there enough of it) as well as the cell quality (are they skewed).  What's Re for the three cases?

[Weiqiang Liu]

Hi,

The flow field looks like reasonable before it fails. The mesh number is large enough and cell quality is perfect. The Re in my case is very small which makes it laminar.

I also found when I further decreased the channel diameter to 1.5 mm, divergence never happened. 

I am really confused.

Best 

Weiqiang

[soto6942]

I am really interested in the subject, so I comment as a UP

[Weiqiang Liu]

Hi,
I solved the problem. It looks like step by step convergence strategy works.

Best

[sufia.khatoon]

Hi Weiqiang Liu,

Can you please explain what is step by step convergence strategy?

I am facing a similar problem. My simulation works for 1.5 mm diameter but there is divergence after few iterations when I reduce the diameter to 1.1 mm and 0.9 mm.