[stefanos]

Dear community,

I am a freshman in CFD and i am planning a fully-developed turbulent (Re~150000), incompressible 3D pipe flow modelling. The pipe is very long and exhibits slight variating curvatures, so slight that they do not intend me to create denser meshes on the curved locations, actually. Due to this slight curvature i am using a Realizable k-ε model, and a 1st order SIMPLE scheme. After a couple of attempts and re-meshes, i have come up with around 1.000.000 elements (the smaller log layer elements have aspect ratio 2/1000) and nr of iterations for the solution to converge (at 10e-5) of around 1200. Still i believe that such large nr's both of elements + iterations could be decreased, though. I would like to ask whether somebody could introduce me a reference on long pipeline turbulence modelling, or a couple of tips on turbulent modelling on such large domains. (I am aware of that my case sounded quite "extreme" and someone could insist simple "Bernoulli" calculations for the pressure-drop, but still there are some effects on wich i am considered....).

Thank you in advance

Best regards

SK