I need some help: Trying to use Fluid Flow

[aparadis]

Okay, I am trying to simulate the flow of air across a intake valve inside of an intake port. I am essentially trying to simulate what a flowbench would be doing in measuring cfm across a pressure differential. The pressure at the port opening would be standard 29.9213 in Hg and 28 in Hg at the outlet of the cylinder bore. I have modeled the air portion of the port with combustion chamber and cylinder bore. I have also gotten a surface mesh and solid mesh done using the bondary layer mesh. I have also place static pressures at the inlet and outlet. The problem that continues to occur has something to do with the load curve and I have yet to find a good explanation of how to properly set it up. Each time I attempt to run a simulation, it shoots back an error saying it cannot find the load curve point. Quite honestly, I'm lost on how to properly set up the load curve. It seems that the static stress simulations are way nicer on the user than the fluid flow.

Model type: 3d

Material: Air

PS: I initially used SolidWorks to get this data but SolidWorks does not break down there velocity results very well. 

Also, thanks for any and all help!

[Joey.X]

Aparadis,

For fluid flow boundary conditions, (pressure inlet and pressure outlet) has more challenge by nature in numerical convergence than the combination of (velocity inlet + pressure outlet) or (velocity inlet + i.o outlet). 

For your specific model, the issue you have may be related to some other reasons. FYI, here isa check list or work around. 

- Check if gravity is disabled

- Check mesh quality and mesh log file for any problems

- Check if there are other boundary conditions making the model over-constrained.

- Tuning up load curve has some help for convergence difficult problem, but not very effective for high-nonlinear fluid problem. Except you really wants the result in under such load curve loadings.  Fluid solver uses internal relaxation scheme for convergence which works better than load curve.

- Change steady state analysis to transient analysis, it helps relief the convergence difficulties.  

- Change to (velocity inlet  + i.i out) boundary conbination if it is reasonable in physical modeling to numercial model setup, this depends on your model physical problem statement.

BTW, if your model is not big, could you upload the whole model so that I can help to have a look?

[Anonymous]

can you upload a screenshot of your analysis parameters?

[Joey.X]

The screenshot of my analysis parameters does not apply to your model, it really depending on the physical model and your assumptions at converting it to numercial model. 

back to your special model in your first inputs, if you stick on pressure inlet and pressure outlet, I suggest to chage original (29.9213 in Hg+  28 in Hg) to (1.9213  in Hg + 0 in Hg), which will make identical model as you know it does not change the pressure difference, but this helps for numerical solution convergence.

[aparadis]

Alright guys, sorry about the long delay. I was out of town and had an engineering exam to prepare for.  I was going to upload the fem file but thats not compatible with the accepted file formats. What type of file you should I send.

I will in mean time upload a view of the analysis parameters.

[Joey.X]

.fem file is not complete model, please do archive/create, it will make a complete model (.ach file). 

[aparadis]

Okay, I converted the file.