Have anyone check if the results from both version are SAME?

It is very difficult, sometime to obtain the same results even by running the same problems on the same software.

amit

I've just ran same cas&dat files on both version and got two different results. What would this mean? Having no changes in the Turb.model and other parameters, shouldn't both versions give same results? Or has FLUENT modified something without telling us?

Hi!I

t depends on what you are simulating. For example multiphase simulations are very different in version 6.1 than 5.5. Some UDf functions also don't work in older versilon and need to be rewritten...

regards

MATEUS

Test case was : Standard k-e and Quick scheme (assuming both versions are same) for a fully pulsed round jet.

I started with older version and UDF was working fine.

Very early stage of the simuation, FLUENT 6 gives me more realistic results.

I guess I can try STEADY JET without using UDF, see if both models give out same results.

Why and what kind of UDF did you use for fully pulsed round jet?

Lola

it's similar to one of FLUENT exercise using sinusoidal velocity, but I also include k_in and e_in profiles at the velocity inlet which are formulated from V_in.

Hi,

i have a prpject about air and its runing and circulation, turbolance in the patient body lung and ventilator set please help me if it is possible for you or guide me how can I take like this information.

your sincerly

hosh

If that's the case, can it be possible that Fluent improved the accuracy and convergence of it's k-eps?

Decide what you want to find out.!!! 2D or 3D, steady or unsteady.. make a list of things you want to find out , this would be the first step I guess,.

That's what I thought.

However, if the model has been modified, they can no longer call it as "standard k-e"????

I haven't managed to find any changes in turb model constants or model ifself from FLUENT Release News.

Am I missing something???

It may be just the case that they've played with the numerical algorithm of k-eps, rather than the model itself. Then they'll be still solving the same standard k-e, but with improved numerics. I guess Fluent user support should have the best and most correct answer to it. However, from the release notes I don't recall any changes to k-eps.