[duaiduaihu]

The problem has bothered me grearly that the results using laminar and turbulent model are so different.

what i am simulating is the motion of the residual water in the oil pipe. the geometry is that the beginning section is the horzontal pipe about 1m, the next section is the upward inclined pipe with the angle of 30degree and 3m long. the following is the sketch:

.................... .......outlet
................ . .......
................. ...
............ . ....
inlet
the initialization is patched with water in the lower part of the horizontal pipe, the inlet is velocity lnlet,and the outlet is the pressure outlet. VOF
menthod is used to model the two phase flow. surface tension and wall adhension are considered.

when the laminar is selected, the phase distribution is similar to the experiment, the interface is wavy. but when k-e model is selected. the water gathers together just likes a huge water droplet,(just like the water in the lotus leaf),what is more ,the interface is very smooth, no matter how large the inlet velocity is .

can any friends give me suggesions. what is the problem with the turbulent model .