Combustion Modelling – Swirl,3D, EDM, Steady-state. Energy residual need help!

[Jee Loong Hee]

Hi,

I have been running a combustion case but the energy equation appeared to be stuck indefinitely at '3e-5' and struggle to go down. Would be very helpful if I could get some insights from you.

This is my problem specification: 

• Mesh 6 million cells
• There are 16 'swirl' holes to inject air and one fuel hole in the middle. 
• Hybrid mesh, predominantly hexahedral mesh + tetra in bridging zone (converted to poly)
• Flame developed (looks very logical shape)
• -Aspect Ratio: Average 5.8142 (below 35 guideline)
  - Orthogonal quality: Average 0.8973
  - Skewness: Average 0.1377 (below 0.33)
• Case ID: 1 (methane-air), Eddy Dissipation Model
• Steady-state simulation, coupled, pseudo time-scale factor 1 (now reduced to 0.8)
• Turbulence: K-e realizable, Standard Wall Function
• Radiation Model: Discrete Ordinates 3,3,2,2, default 10 interval
• Gas absorption: wsggm-domain-based
• Currently switched DO and Energy to 2nd order.
• Gradient scheme: node-based 
• Shape of furnace, very uniform, burner in the centre of a rear square face and extruded square face that minimic a rectangular furnace, then the rectangular face entering a pyramid shape converging section and then exit via a rectangular slot. 
• Total iteration achieved on 67000 iterations. 
• Heat of Reaction source around 65kW now (according to the experimentalist, it should be higher based on fuel and air value supplied)
• Net result (W): 15242 ! (can't find a way to get it down to 1 - 2% error). That explain why the energy is e-5. For other projects, I have reached -7 to -9 before
  
  I attached a snapshot of the temperature field to show you that my flame is developed (it is only a quadrant). I can't understand why is this burner with swirl hole so much harder to get lower energy compared to my past projects (different burners)? I also attached a convergence monitoring plot to show you that I have been very careful with the simulation progress.
  
  
  
  
  
  Would be very helpful to hear from you, how I can bring my simulation to 'next level'.
  
  Thanks again.
  J
  
  
  
  
  

[Jee Loong Hee]

Would be very helpful, if I can speak with ANSYS expert directly on customer portal too via file sharing. That would be greatly appreciated.

[Rob]

If you've got commercial TECs please log a ticket, otherwise, it's the Forum. 

That looks like the flame region is moving slightly with iteration, so you've likely got some transients in the results.