Now I meet a problem confused me. I establish a blockage(0.2m x 0.2m x 0.07m)that top surface has high temp. glass inflow and bottom surface has about 1200 small holes(each hole diameter is less than 0.002m)to outflow. When I finish my model and run it. The program shut down. One reason may be my computer ram(256MB)lack . Another reason may be holes too small that will grown too many small grid region. So who can tell me "Phoenics can run a model in that has small size model"?

Brian,

I ran into a similar problem. For me Phoencis could not handle processing the shear number of grid points and elements in the model. The problem was not the size of 'holes'.

So for you problem i do not think that it is the size of the holes that is the problem, but rather the amount of data the pheonics needs to process. Have you considered using BFC, and using symetry in the design to reduce the shear number of holes in the model?

I hope this helps.

Leon

Dear Leon

Thanks your suggestion. For my case all outflow holes(1200 pcs)is 2 dimension and is located on a plane. So It do not use BFC and I have used symmetry model. IF I wanted to see the flow condition of holes, I need to make the hole(2 mm) prodce many grid such as partition 10 . It will make every grid size less than 0.2 mm. For my case it make phoencis program shut down. I will appricate your respondence.

Brian,

From mmory there is something called an F-Array in phoenics and it has a certian maximum size. Once you have reached it, phoenics shuts down. The upper bound maybe altered, but I have not done this so I cont say for sure. Try contacting CHAM direct and asking about tips on reducing the model size.

I my problem I reassesed the accuracy I required with my model and sacrified the accuracy of the solution and the model to obtain some data.

Sorry I could not be of any further assistance on the matter.

Leon

I am not sure the reason of your problem, but i think the RESULT file may be helpful to you.

Here at Flowsolve we use a different approach for this sort of problem. There is no point in trying to include each individual hole in the model. Even if the model is not excessively big, unless your grid size is considerably smaller than the hole size you will not simulate the flow through the holes with any accuracy.

It is better to adopt a more pragmatic approach. Ask yourself "What is the main effect (or effects) of the holes that I wish to include in the model?" If (for example) it is the pressure drop, corresponding to flow resistance of the holes, a better approach is to do the following.

* Don't include each individual hole in the model. * Imagine instead that the effect of the holes is spread out, so that you can use a coarser grid. * Apply a suitable correlation for pressure drop, based on hole spacing, diameter, fluid properties, etc. Such formulae can be found in fluid dynamics / heat transfer textbooks. For example, a resistance formula of the type force = const * density * (velocity)**2 may be appropriate.

Remember - always start with a simple model on a relatively coarse grid, then refine step by step. That way if the model breaks, you know what broke it.

At Flowsolve we find this to be a very effective approach.

Good luck!