Hi all,

I am modeling a toiletbowl. In setting up the model, I have a few doubts whether or not my approach is correct.

1. Is it necessary to use 'free surface' option to model the water flow in and the remaining water in the bowl? Or using two materials (one is air and the other is water) is sufficient already? in other words, without using the free surface option, will starcd know that water and air in the toiletbowl, ideally, does not mix and react?

2. as for the boundaries, i am using inlet and pressure boundaries. inlet is self explanatory (i think!). the 'static' pressure boundaries are applied at both, the outlet and the open area (where we sit). now, how do i set the pressure at the outlet that initially it is the same as atmosphere but later, it changes?

i ran many runs already and yet the solution always diverges. i played around with the time steps (i believe it is small enough), under-relaxation factor, # of sweep and relative tolerance and yet, no change. so please help. Thanks in advance!!!

shazlan

The answer to 1) is yes, you have to use free surface modelling. This is exactly what it is intended for.

Your boundary conditions are a little trickier - I would use inlet for the injected water, pressure for the top of the bowl, and pressure for the outlet in the U-bend.

If you use the static option rather than piezometric for this pressure boundary, there is no need to allow for the water pressure due to gravity, it'll be added automatically.

There is no reason I can think of that this setup would give trouble, provided your time step is small enough. It must, of course, be a transient solution, so there is little point in messing with relaxation parameters and inner sweeps.

Dear Brian,

Thanks for your helpful comment. However, I still have a few more questions, if you or anybody else can help clarify it to me.

1. Yes, for the top of the bowl, I am using static and also, enviromental option. Hence when you said that there's no need to allow for water pressure due to gravity, do you mean just leave the pressure value box in the pressure BC as 0?

2. If not, how do I define the direction of the pressure? If I set the parameter to be -101325 Pa (atmospheric), how is this be different from 101325 Pa (without negative sign)?

3. In the scalar boundaries, do I need to define another scalar (H20_l) to let Star-CD knows that the flow coming in is liquid water? Or can I simply use the already define VOF? Also, just as a point for clarification, by setting the mass fraction, of the first inlet for instance, to 1, it means that the inlet flow consist of nothing else but the defined scalar, right?

I really appreciate the comments, hints and help from you guys. Thanks and have a nice day!

Shazlan

Besides that, I have run the analysis a few times and everytime it reaches iteration #19, it takes forever to get to the next iteration - so far, the longest i've waited is about 12 hours. Is this normal or there's something going on?

Shazlan

HI! Shazlan

This simulation project has beem done before in Taiwan. In this project there are nothing special need to set up beside the count number.

And the other thing need to notice is the big pstt file need to seprate.

tomosada

Ok, I'm not aware of that. What do you mean by the 'count' number'? Is it the number of iterations? and yes, the pstt files are separated - for mine, each is for 1 secs of sim time.

Anyhow, do you have any reference, online link maybe, on the project you mentioned above? I would like to see if there's anything wrong or anything I can improve on with what I've been working so far.

Thanks!