Water flow through cylinder with one inlet and five outlets


I have gone through the relevant tutorials and searched the articles available here for help, but still can't seem to get my simulation to work. I will refer to Figures in this post, and all of them can be found here: https://www.evernote.com/l/AkEGNoMj4VdGd6OaDcooC3q_Qo6HejB6zZs

-What I want to simulate is water flow through a cylinder with one inlet on one face and five outlets on the opposing face. See Figure 1. The simulation is transient in the sense that it is supposed to simulate the flow that occurs when a syringe is connected to the inlet, and pressing the syringe will send a small amount of water flowing through the cylinder. I want to be able to simulate the impact different orientations in reference to the direction of gravity.

Question 1. Should my substance in this case be 'Water' or 'Air and water'?

Question 2. Should the simulation be of a transient state or a steady state?

Question 3. The option to set g is not available when I use just 'Water' as substance. I assume this means that the answer to question 1 must be 'Air + water'? Or is there some other way to take gravity into account?

-After creating all the cylinders and positioning them correctly, I united all of them to a single flow volume.

-I then specified that all faces in the geometry be 'Wall' conditions. Next I set the inlet and the five outlets to have the 'Inlet' and 'Outlet' conditions, respectively.

-For setting the values of initial conditions and reference values, I followed this tutorial by and large: www.symscape.com/node/818
My cylinder is aligned with the z-axis, so I set the initial conditions to be -2 m/s flow from the inlet into the cylinder.

Question 4. What is the purpose of the Substance-reference properties (such as velocity reference)? I set my velocity reference to 0, and assumed the initial conditions are all I need, but I need some clarification here, since I feel the tutorial told me what to do, but not why to do it.

-I kept following then went on to the result and dragged a velocity vector field onto the flow volume. I tried to visualise it using both arrows (Figure 2) and a colour map, but only the arrow options seem to work, even though I follow the exact same steps in both cases. The colour map doesn't show up at all.

Question 5. Any idea what's going on here, with the color maps not working?

-I ran the simulation and did get some results, shown in Figure 3.

Question 6. Do these look realistic at all? I don't know much about CFD, but the back-flow doesn't look that realistic to my untrained eyes at least. What can I have done wrong in this simulation and in general in building this whole model?

Thanks in advance, I look forward to your replies!

Mikael Eriksson

No votes yet

Start simple single-phase (water) steady-state

Assuming the chamber is already full of water and that the chamber diameter is of the same order of magnitude as a standard syringe I doubt you'll see any variations due to gravity. Gravity will have an effect in a multiphase flow for air and water, i.e., if the chamber initially was full of air and you are injecting water.

I would suggest you start simple using a water substance and steady state. As you have an axis-symmetric flow volume, I'd also suggest you simulate a quadrant as in the tutorial.

I doubt your current simulation setup is valid, given you are using multiphase and unsteady. It sounds like your flow volume is valid, given it's a single water-tight volume.

Is this flow likely to be laminar? If so you should set Substance->Solver->Turbulence Model = Laminar

The reference velocity needs to represent a characteristic velocity, e.g., the inlet velocity. Then it can be used as the inlet velocity and the initial velocity. You can override it within the inlet and initial condition options, but you need to set it some realistic, not zero.

I also recommend that you create a Residual Monitor so you can determine when or if your simulation is converged, as we do for all the RANS Flow tutorials.

I'm not sure why you aren't seeing a colormap, it is likely caused by a problem with your simulation. If you follow my suggestion for a simplified simulation similar to the tutorial you are using as a template then I'm confident the colormap will work.

Thanks for the

Thanks for the reply!

Actually, the chamber is not already full of water. I am indeed investigating what happens when injecting water. How does this change your suggested ideas?

Yes, I believe the flow is likely to be laminar. Did some estimate calculations that indicates so.

Got it about the reference velocity and using residual monitors! I actually did use residual monitors when I tried to simulate using the 'Air + water' method, and saw that the solver stopped after the first time-step due to running into some singularities.

I'll try again to do a simulation using a steady state and water substance just to get some validation.

Could you please specify how I can then get the transient injection simulation set up?


Water and air multiphase simulation

Once you have successfully simulated the simplified (water only) model, review the tutorial "Transient Water and Air Flow in a Tank" with particular attention to the setup of the substance, boundary conditions, and initial conditions for your multiphase unsteady version. Note that unsteady multiphase simulations typically require fine meshes and take a long time to run.