Premixed Combustion at Higher Swirl Numbers

  • 26 Views
  • Last Post 26 February 2019
krany3 posted this 19 February 2019

I'm analyzing the premixed combustion at different swirl numbers using Realizable k-epsilon model for turbulence modeling. As the swirl number increases the flame will move upstream. But, when the swirl number is greater than 0.6, the flame moves farther downstream. Is it because of the k-epsilon model? Below image is the temperature contour.

 

Order By: Standard | Newest | Votes
rwoolhou posted this 20 February 2019

Possibly, for high swirl you may need RSM for turbulence. You also need to review the flow field incase there's a change in the flow structure and/or flame stability mechanism. 

krany3 posted this 20 February 2019

I checked the flow field. It's not symmetric as it's in the lower swirl numbers. Even I tried with RSM, starting with realizable k-epsilon model as an initial solution. But, I'm getting the similar flow field. Do I have to change the under-relaxation factors? 

Thanks

rwoolhou posted this 20 February 2019

Sounds like you're getting a core forming: you're going to have to look at the flow in more detail. The core may also be trying to move. Have a look at the theory as you're moving into an interesting flow field study. 

krany3 posted this 20 February 2019

What do you mean by core formation? I did not get you.

Thanks

krany3 posted this 22 February 2019

I checked the flow field with RSM model. But, the flow field is not all detected by RSM. Still the temperature contour looks the same as in realizable k-epsilon model. What could be the reason?

Thanks

rwoolhou posted this 22 February 2019

Swirl combustors stabilise the flame using swirl to reduce the axial velocity to below the flame speed (to paraphrase one or two books). You then have what's called a vortex core (think hurricane) that's never stable, and rarely symmetric. 

krany3 posted this 22 February 2019

But, I'm not getting the flow field at all in RSM. What could be the reason?

krany3 posted this 24 February 2019

I included a rod in the injection tube and used realizable k-epsilon model for this same case with the swirl number greater than 6. This time it predicted the flow field properly. 

krany3 posted this 26 February 2019

The courant number was very high for the model. So, I decreased. The model predicted the flow field correctly. Does courant number affect combustion?

Thanks

rwoolhou posted this 26 February 2019

It'll affect solver accuracy & stability, and that will affect combustion. 

Close