CFD Questions

[ArthurPain]

All,

I could use some guidance re: CFDing.

For diffuser analysis:

How fine a mesh is required? Is a finer mesh around the restriction a good idea?
What boundary conditions are used? Pressure drop? Vol flow at a given speed. Exact numbers would be helpful.
What metric do you use for comparison? Pressure drop? Mass flow rate?

Also, what is the pest way to evaluate the air box/runner design? To decrease complexity of the model can you remove the diffuser from the analysis and replace it with the boundary conditions imposed with the diffuser in place?

Tnx, AP

[ArthurPain]

All,

I could use some guidance re: CFDing.

For diffuser analysis:

How fine a mesh is required? Is a finer mesh around the restriction a good idea?
What boundary conditions are used? Pressure drop? Vol flow at a given speed. Exact numbers would be helpful.
What metric do you use for comparison? Pressure drop? Mass flow rate?

Also, what is the pest way to evaluate the air box/runner design? To decrease complexity of the model can you remove the diffuser from the analysis and replace it with the boundary conditions imposed with the diffuser in place?

Tnx, AP

[BeaverGuy]

Generally you impose a pressure drop across the system. Most of the things I have seen use a 20kPa drop. I personally compare flow rates when evaluating a design because that is what I'm interested in. Power is proportional to flow rate.

For the airbox and runner analysis if you can use the exit output values for your restrictor analysis as inlet boundry conditions for the airbox, that would best. However, you would need to set up both tests so that they were essentially operating under the same conditions.

[Marc Jaxa-Rozen]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Generally you impose a pressure drop across the system. Most of the things I have seen use a 20kPa drop. I personally compare flow rates when evaluating a design because that is what I'm interested in. Power is proportional to flow rate.

For the airbox and runner analysis if you can use the exit output values for your restrictor analysis as inlet boundry conditions for the airbox, that would best. However, you would need to set up both tests so that they were essentially operating under the same conditions. <HR></BLOCKQUOTE>

This is also how I do it. I used to mesh the entire intake system, but setting up a boundary for the restrictor exit instead of modelling the whole thing saves a bit of computing time and doesn't affect results.

Does anyone have experience with coupling WAVE and CFD? We've been toying with WaveMesher to convert 3D geometry to WAVE's 1D engine, but the results don't correlate with Fluent for more exotic plenum shapes.

Marc Jaxa-Rozen
École Polytechnique de Montréal

[woollymoof]

I'd be careful just running the diffuser in the analysis and assuming the results will correlate with the whole intake. It depends on your intake but I have seen vastly different flow patterns between just running the diffuser and running the whole thing.