In modeling a gas flow in a pipeline using fluent, I have used steady gas flow with velocity inlet as inlet boundary condition and outflow as outlet boundary condition. I have modeled this and The result outlet pressure is 200psi, as was expected, but the problem is in the second part of the modeling which is an unsteady state in which a valve will shut off the flow inlet of the pipe, and there will be no more gas flow entering the pipe. The outlet pressure is expected to decrease gradually as the pipeline is discharging. I want to model this unsteady condition and have the outlet pressure change vs time. I have tried many methods but there were no acceptable solution to this problem. Eg I read the steady state data in unsteady case, to use it as initial condition of unsteady problem, and I have used mass flow inlet=0 as inlet boundary condition and outflow as outlet B.C of the unsteady case and run the program, but it seems that fluent does not consider the initial condition. How can I model this condition in fluent?
I have used dynamic mesh too. I modeled it like an example in the tutorial of fluent help. but the problem is that, when I use velocity inlet boundary condition the flow enters the pipe and strike to the valve which is in front of the pipe, and due to this after running the steady state case in the fluent, the velocity has a large reduction after striking the valve and the pressure increases significantly. so the result of steady case becomes unacceptable.
It is referred in the Fluent User Guide that outflow boundary condition cannot be used in case of unsteady flows (Fluent User Guide 6.3.10.), you can try other boundary type or think to replace outlet with any system could maintain your goal with considering the mass conservation principle.
Solve the steady state situation. Then switch to transient mode. Run it for a few time steps. Now write a transient table for velocity inlet. Use it as boundary condition such that the velocity reduces from the initial value to zero within certain time resembling the actual condition. And then run it further.
initial condition in unsteady flow problem is defined in the "manu box" of the solver of the FLUENT.
Abdul Rahman Mallah: the reason that makes fluent help tutorial not to use outflow boundary condition, is that the fluid is compressible. in fluent the outflow boundary condition can not be used in compressible flows.but in my problem the flow is incompressible and I do not have any other condition to use.the thing that is obvious to me is that this wrong result is not due to outflow boundary condition, but it is due to velocity reduction after the valve.
Viraj Shirodkar & K. Hami: I have tried this method too. I model the steady condition without considering valve and dynamic mesh.in steady state condition the results are correct.then switch to unsteady state condition and use a velocity profile that is variable with time.and the velocity reduces from the initial value to zero in a short time.I used a step function in the udf, in which the velocity has some different values in some times.but there is not good answers.and when the velocity decrease during time, the pressure increases unlike what I expect. it seems that fluent can't distinguish that it the flow exist in the pipe and amount of flow decreases little by little. it seems that it consider only a constant amount of the flow with some different velocities.
R. C. Mehta · could you please explain more about "manu box" of the solver of the FLUENT. how can I find it in fluent? I really be thankful if you explain what exactly I should do.
Yes, outflow boundary cannot be used for compressible flow, and also cannot be used for the unsteady problems (even if the flow is incompressible).
As the Fluent User Guide, outflow Boundary can be used in case of zero diffusion flux for all flow variables. In simple words "zero change of velocity over the time at the outlet", so the outflow boundary is proper for fully developed flows.
Hence, in case of unsteady flow, the conservation laws of mass and momentum in the system will be violated: outflow boundary cannot respond to any change of velocity magnitude at the inlet or at any point of flow field and that will lead to solver divergence.
In your problem, velocity is dropping till 0, so both mass flow and momentum at the inlet will drop to 0, but outflow boundary cannot balance that drop since outflow should maintain zero diffusion flux for all flow variable at the outlet, That leads to solver divergence.
Try to use pressure outlet with reasonable pressure.
Abdul Rahman Mallah: thanks for your help. my question now is that, the outlet pressure is variable now.and I can't give the outlet a constant amount. is it possible to have a varying pressure outlet in pressure outlet boundary condition?if not, what other outlet condition could I use in my problem, regarding that all variables such as pressure,temperature,velocity etc are variable in the outlet and I don't have any information about their pattern and in fact my goal is to observe this changes as the result.
Could you please tell me in actual, what this pipe is connected to? is it reservoir or atmospheric pressure? or another piping system?
The impact of the pipe network on the behavior of gas at the pipe outlet cannot be neglected, realistic results at the outlet could be achieved by taking into account a portion of the pipe network in the modeling, no need to study the whole network, only the branch where the studied pipe is connected with. In this case, system will have new outlet.
You can select pressure-outlet for the new outlet (pressure value can be estimated by analytical hydraulic calculations).
Original outlet will be part of the fluid domain (no boundary condition to be selected), so fluctuation and variation of the pressure value could be studied there.
as mentioned in fluent guide tutorial the absolute pressure will be negative after the valve is closed and it indicate cavitating flow. I can't understand the meaning of it and couldn't find a sensible meaning for it.what does it mean?aren't there any pressure decrease due to inlet shut off?
Abdul Rahman Mallah: could you please tell me how can I see the real flow pressure?
Hi, If your inlet boundary condition is 0 mass flow rate then it means the velocity at the inlet is 0 ie., there is no convective term. Its obvious that Fluent is not considering the inlet condition. If mass flow rate is taken as inlet then it should have finite number. It is one of the reasons. Try to understand the physics of the flow and then proceed with Fluent.
- Badges
- Science topic