[faraz92]

Hi

I want to model a multiphase flow(solid-liquid) in a stone trap,i have problem with reverse flow,when fluent starts iteration,immediately appear a massage that says(revered flow in ... faces) can any body please help me to solve this problem?

the goal is achieving contours for velocity difference(velocity magnitude),pressure and distribution of solid phase

here are the problem conditions:

laminar flow-multiphase(mixture)-phase1(crude oil),phase2(stone&sand)-granular with 0.005 diameter for stone &sand-velocity inlet boundary condition for inlet of flow and pressure outlet for outlet.there is gravity-and the operating pressure is 20bar-the inlet velocity is 0.000347(m/s)

the geometry of problem has been attached

it should be noted that as you can see in picture there are 125 slots in the inner pipe

[ComputerGuy]

The explanation below assumes you've defined your exit properly as a pressure outlet (which it appears you have). If you've defined it as a pressure inlet, which is perfectly valid, you'll always have this error as fluent is calculating a negative mass flow. A negative mass flow in or a positive mass flow out are the same.

Reverse flow isn't a critical error. Think about pouring water/beer/soda out of a bottle: even though the general trend is for all the water to exit the bottle, there are periodic "back flows" of air into the volume to make up for lost volume. Your simulation, either naturally or through a poor initialization condition, may have backflow in early iterations/time steps as the solver "works out" your bad guess.

If your area of interest will be disturbed by the backflow, extend your exit domain so that the backflow doesn't influence your solution.

If you *know* the backflow condition (which, presumably, is pure oil), simply set the back flow volume fraction of your stones to 0. Let the solution converge. If it converges deeply, you may simply have backflow. Again, not a problem. If seeing the message bothers you, in the text interface type: solve (hit enter), set (hit enter), flow-warnings (hit enter), no (hit enter). This will turn the message off.

Bottom line: let the solver work for a while. If it still shows backflow, consider growing your geometry so the backflow doesn't change your answer.

ComputerGuy

Originally Posted by faraz92

Hi

I want to model a multiphase flow(solid-liquid) in a stone trap,i have problem with reverse flow,when fluent starts iteration,immediately appear a massage that says(revered flow in ... faces) can any body please help me to solve this problem?

the goal is achieving contours for velocity difference(velocity magnitude),pressure and distribution of solid phase

here are the problem conditions:

laminar flow-multiphase(mixture)-phase1(crude oil),phase2(stone&sand)-granular with 0.005 diameter for stone &sand-velocity inlet boundary condition for inlet of flow and pressure outlet for outlet.there is gravity-and the operating pressure is 20bar-the inlet velocity is 0.000347(m/s)

the geometry of problem has been attached

it should be noted that as you can see in picture there are 125 slots in the inner pipe

[faraz92]

Originally Posted by ComputerGuy

The explanation below assumes you've defined your exit properly as a pressure outlet (which it appears you have). If you've defined it as a pressure inlet, which is perfectly valid, you'll always have this error as fluent is calculating a negative mass flow. A negative mass flow in or a positive mass flow out are the same.

Reverse flow isn't a critical error. Think about pouring water/beer/soda out of a bottle: even though the general trend is for all the water to exit the bottle, there are periodic "back flows" of air into the volume to make up for lost volume. Your simulation, either naturally or through a poor initialization condition, may have backflow in early iterations/time steps as the solver "works out" your bad guess.

If your area of interest will be disturbed by the backflow, extend your exit domain so that the backflow doesn't influence your solution.

If you *know* the backflow condition (which, presumably, is pure oil), simply set the back flow volume fraction of your stones to 0. Let the solution converge. If it converges deeply, you may simply have backflow. Again, not a problem. If seeing the message bothers you, in the text interface type: solve (hit enter), set (hit enter), flow-warnings (hit enter), no (hit enter). This will turn the message off.

Bottom line: let the solver work for a while. If it still shows backflow, consider growing your geometry so the backflow doesn't change your answer.

ComputerGuy

thank you very much for your guidance,but the main problem is that the solution has been diverged.

thanks for your help