[FengWu]

I want to used the Flunt software to simulate the flow in the tube(such as Pipelines and Elbows ). Is it possible to get the accurate resistance of the tube?

[pete]

What is it that you want to use CFD to do?

You don't need Fluent to estimate the resistance in a long tube or in a tube elbow. Simple analytical formulas does that just as well and with much less risk of errors. If you want to do a detailed aerodynamic design of, for example, a tube elbow in order to minimize resistance you might want to use CFD to compare different designs.

[FengWu]

Originally Posted by pete

What is it that you want to use CFD to do?

You don't need Fluent to estimate the resistance in a long tube or in a tube elbow. Simple analytical formulas does that just as well and with much less risk of errors. If you want to do a detailed aerodynamic design of, for example, a tube elbow in order to minimize resistance you might want to use CFD to compare different designs.

I want to get the accuarate resistance of the tube, however, when I used fluent to compute the resistance of the tube, I foud that the result I obtained is far less than the value obtained by the empirical formula，especially for the local resistance of the Elbow. If I changed the "Roughness height" and "Roughness constant" in boundary condition, the situation ameliorates, But I don't know what is the exact value for "Roughness height" and "Roughness constant" . For example, if the resistance coefficient is 0.015, what are the "Roughness height" and "Roughness constant" value?
Thanks

[pete]

I'd trust an empirical correlation much more than CFD in his case.

[meb]

CFD is a powerful tool. If wisely used allows to predict pressure drop within an error less than 5%.
We have found excellent correlations for pressure drop in reed valves; I have also simulated an air-air exchanger, also in this case a very good correlation between experiments and CFD has been observed.
As noticed, for standard components it is better to use literature data, CFD helps for design of new components because allows to understand how performances can be improved; this means that you can go well beyond the pressure drop observed experimentally: you can see how (and where) the losses are generated.