Combustion: pool fire problem
 

Hi,

I am using the Eddy Dissipation Model in CFX-5.6 to simulate a burning pool fire (methane/air mix, 1.0m x 1.8m rectangular source in a small room). I set a mass flow of 0.08 kg/s on the fire inlet, corresponding to a fire size of about 3.5 MW.

With time steps of 0.25s and 10 iterations per step, I got a good looking plume up to about 7.5s, which then moved across to one side of the pan whilst thinning out, subsequently disappearing. Seems like something stopped the combustion process for some reason. Velocities around the pan are quite small, so it's not like having a wind blowing into the fire.

Has anyone else had a similar problem modelling pool fires and how was it resolved? At the moment I am trying smaller time steps (0.1 s) but keeping a similar number of iterations per step in order to keep the problem manageable.

Jon.

Re: Combustion: pool fire problem
 

Could be something to do with flow towards the fire blowing it out. Anyone had a similar problem?

Re: Combustion: pool fire problem
 

Hi,

I think you will need to hit the textbooks before getting too deep into this simulation. A pool fire is a very strongly coupled probem by which the heat release of the fire drives the heat transfer, mass transfer and flow processes.

The eddy dissipation model is a drastic simplification of the entire combustion process. It is an assumption that the only important process is the time scale of trubulent eddies to "dissipate" ie mix beyond their distinct identity. Since combustion can only occur when fuel and oxidant are mixed down to the mollecular level this model assumes that this is the only process of relevance.

I think you will need to take a few steps back on this one. There is not much use in playing with time step size and number of inner iterations unless you have some relevant physics.

A good place to start is some standard references on poll fires such as:

Introduction to Fire Dynamics Dougal Drysdale

There are also quite a few publications on pool fire simulations in the literature.

Best of luck,

Bak_Flow