Learning Forum

[Weiqiang Liu]

Hi all,

I posted some threads before for a soot cake combustion project. Now I get some results. The maximum temperature during combustion is very close to literature and soot is consumed in similar duration with literature. At first, I believed that I just got perfect results. Then I found the shape of high temperature region was different with literature though the magnitude was very close. I put literature and my result for temperature contour below:

The maximum temperature is very similar which are 853k and 858k respectively. It's just the shape of the high temperature region is different. Then I found that the inconsistency of temperature distribution is because I have very unreasonable velocity field.

I checked my Y-velocity contour and found my model could not capture the effect of end wall. There should be a very large negative Y-velocity at end of the channel wall and gas was forced to flow through porous media to the outlet channel.

I reloaded my UDFs and UDS and then calculated a pure flow instead. Perfect velocity contour was obtained as follows:

As long as UDFs and UDS are involved, velocity contour becomes very unreasonable. More strange thing is though my model can not capture effect of end wall, it can capture the effect of a porous step in front part of the inlet channel. I put the Y-velocity contour near this step below:

At first, I suspected it was due to continuity equation was not converged. Then I checked the pure mass flow between inlet and outlet. The net mass flow showed that continuity was satisfied.

Can anybody give me some suggestions?

Best 

Weiqiang

[Amine Ben Hadj Ali]

So without UDF looks good? Even with porous zone right? Gravity us on? Do the UDFs affect the density or viscosity somehow?

[Weiqiang Liu]

Hi, Amine,

Yes, without UDF, velocity contour looks perfect with porous zone. Gravity is off. I have a temperature dependent viscosity UDF in source code. 

Best.

Weiqiang

[Amine Ben Hadj Ali]

So the transport is influenced.. Can you check the values of the viscosity if they make a sense.?

[Weiqiang Liu]

Yes, I checked the value of viscosity.  I used temperature- dependent equation which is from DIPPR database. Viscosity under 523K is 2.6715e-5 pa.s 

I think the viscosity value makes sense.

Best 

Weiqiang

[Rob]

Please can you replot with the node values off? Also can you plot the viscosity in Fluent to check there's not an error with the temperature dependency. 

[Weiqiang Liu]

Hi, 

I found it's problem of permeability. If I give default value of permeability in fluent, continuity equation converges very well and similar velocity contour with literature can be obtained. However, with this very high default permeability, pressure drop across the porous media is totally unreasonable. If I just set normal soot cake permeability, pressure drop is normal and velocity contour becomes unreasonable again.

Therefore I guess, the literature is wrong

Best

Weiqiang

[Amine Ben Hadj Ali]

Thanks for checking that. Be careful when setting that in Fluent: Fluent expects you are setting the inverse of permeability (1/alpha) not not alpha and not mu/alpha.

[Weiqiang Liu]

Hi Amine,

Yes, I give 1/alpha in fluent. Thanks very much for your help.

Best 

Weiqiang