Fluid flow through a nozzle

[jdoggk40]

Hello.

I am attempting to perform an unsteady fluid flow analysis on a nozzle in Algor, but continue to get convergence errors. The nozzle has a .25 inch inlet on the top. The outlet is a 3.6" X .03" outlet on the bottom. I set the inlet at .02 m/s and the outlet is set as a prescribed inlet/outlet.

I've tried setting the load curve up in several different configurations (0, 2, 4, and 6 sec increments with 3 steps per time step and a multiplier of 1....also tried starting w/ a multiplier of 0 as recommended by algor, etc...).

Algor gives me the error: "The solution did not meet convergence criteria, try the following" and lists "1) reduce the loading, 2) relax the convergence criteria, 3) increase the # of interations, 4)change the type of norm (IRE=0 or 1)"

Does anyone know of any documention regarding fluid flow analysis or have any recommendations on time steps etc.?

Thanks in advance,
Jesse

 

[GBor]

I'm not a fluid flow wizard, but for unsteady flow, that doesn't seem like nearly enough steps per time unit. Even at 0.02m/s, you are looking at it traveling 0.007 meters (or 7 mm) which is a pretty good distance for a nozzle that is only 3.6" long. The other comment I would have would be to watch your units. You've cited inches and meters in the problem statement.

To capture better behavior, I may try as many as 100 steps per second, but at a minimum, I would give 10 a try.

Garland

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group
Magnitude The Finite Element Analysis Magazine for the Engineering Community

 

[zcp]

The Algor help files give good guidance for setting these parameters, even to the point of equations based on fluid velocity and size of model. There are several solvers that you occasionally have to try back and forth (see the help files for good descriptions of when to use each). You can also try slowing things down to make sure your model works or to test out different solvers and parameters. Watch your transitions and edges in your model and definitely "check those units!".

ZCP

http://www.phoenix-engineer.com

 

[jdoggk40]

Thanks for the advice! I got the analysis to run shortly after posting. It turned out to be a problem with the mesh, as the outlet was extremely narrow. I added mesh refinement points to ensure two elements covered the width of the outlet. The analysis ran with no problem following this change.

Thanks again, Jesse