Hi, everyone,

I am solving the mixing of gas-liquid system in a small reactor. In my experimental work, I found the diameter of some gas bubbles was about 1.5 mm, however, I divided my reactor into thousands of cells, whose dimension was 1.0 mm, which means that the diameter of gas bubbles from CFD calculation are different from my experiments. I think the gas volume fraction will be different too.

Anyone has some ideas about how to solve this problem?

thanks in advance.

I am just curious, could you tell me what cfd code are you using ?

Thank you

Danny

Hi,

I am using CFX4.1. You know, I intended to solve the problem with a small reactor ( the diameter is 0.008 m ) and I had divided the reactor into thousands of cells ( the dimension is about 0.0002 m), however, the gas bubble diameter may be greater than the size of a cell, which means you can not solve the gas volume fraction in a so small cell. That is my problem.

Tiger

Hi,

May I suggest you to use CFX discussion forum. I believe someone over there can help you.

Danny

The disperse phase models in CFX-4 are based on a source term in the momentum equation which couples the two phases and represents the drag force on the bubbles at the slip velocity (Continuous Phase Velocity-Disperse Phase Velocity).

There is no need to resolve individual bubbles because the model inherently assumes that you are not, and that the bubbles themselves behave as if they are a single fluid. The only reason you should need to worry about the bubble diameter is it's effect on the source term coefficient, or drag coefficient in this case.

Hi, Dan Williams,

You are right, in my case, how to calculate the drag cofficient is my problem. Do you have any idea about it?

Tiger

If you enter the particle diameter and you are running the continuous-disperse model, then you have a selection of models for the drag coefficient. I can't remember the exact formulation for the source term off the top of my head but it is something like: 1/2*bulkdensity*Cd*particlediameter^2*Uslip^2/(ralpha*rbeta) (this is just a guess). There are several correleations for Cd in this case. One is Schiller Naumann I beleive. There are others that might be more appropriate to your application. Read the section on multiphase flow in the CFX-4 manual.

Dan.

Hello, Dan,

I hope my problem is that the diameter is changing from point to point, so I have to calculate the bubble diameter at first, however, the diameter of the bubble mayber greater than the dimension of the control cell. If I can get a reasonable bubble diameter, so all of the problem will be solved.