Fluids

Should the flow be compressible to study combustion dynamics in a combustor?

• Akash564
Subscriber
I'm trying to study combustion dynamics in a combustor model. Mach number is subsonic and very small. Should I use compressible flow to model combustion dynamics? n
• Rob
Ansys Employee
Does the gas expand when it gets hot? n
• Akash564
Subscriber
Yes! The gas expands after combustion.n
• Rob
Ansys Employee
Then you may want to consider compressible flow. Depending on the pressure loss incompressible ideal gas may be sufficient. n
• Akash564
Subscriber
I've used incompressible ideal gas for density calculation and I did not see any changes in the pressure inside the combustor. Should I use ideal gas for density calculation? How do I implement compressible flow? nThanks!n
• Rob
Ansys Employee
Depending on the system you may not see much change in pressure as the gas exits the system. It depends where the pressure drop in the system actually is. nCompressible flow means the density changes with stuff ie temperature and pressure, there's nothing overly complicated about the set up. This does change a little when you start looking at much higher speeds (shock waves etc) but if you're getting those in a combustor it'll be an interesting design. n
• Akash564
Subscriber
I get it. How do I model compressible flow in Fluent? Is there any better non-premixed combustion model to study combustion dynamics?nThanks!n
• Akash564
Subscriber
Should I use the density based solver for compressible flow, even if the Mach number is low?n
• Rob
Ansys Employee
Compressible flow means using a temperature & pressure dependent density: I'd suggest ideal gas for most applications. You then need to match the combustion model to the physics. Given we don't even know what you're burning and under what conditions we can't comment. n
• Akash564
Subscriber
Thanks Rob! I'm modeling non-premixed combustion at atmospheric conditions with methane-air mixture using FGM technique. I don't have an option to select ideal gas for density with flamelet mixture. It is incompressible ideal gas for the mixture. I don't see combustion with both FGM technique and non-premixed combustion models. n
• Akash564
Subscriber
Hi Rob! I've tried to model combustion with both FGM and non-premixed models. I don't see combustion with both the models. Can you please help me with this?n
• Rob
Ansys Employee
If it's not combusting check if the AFR is sensible and that you've got an ignition effect in the domain. Some of the models require a small amount of product when you patch: it'll be noted in the manuals. n
• Akash564
Subscriber
I did patch with reaction progress variable for FGM. Still I don't see any combustion. Initially there was irregular combustion in the domain without boundary layers. But, when I introduced the boundary layers, the mesh was changed and no combustion was observed. Is this because of the change in the mesh?n
• Rob
Ansys Employee
Possibly, but the new mesh could mean the flow is more correct and that the combustor isn't suitable for stable combustion. Where did you get the chamber design and air & fuel boundary conditions from?n
• Akash564
Subscriber
I'm testing a new combustor design at different air and fuel boundary conditions. We have run some cold flow experiments at our lab and I've used the same swirler geometry and operating conditions. Will the flame stability depend on the size of the confinement?nThanksn
• Rob
Ansys Employee
What confinement? Most burners require air in the right amounts in the right places, and also that the stability mechanism is matched. Have a look for work by Brundish and Wilson from the late 90's on fluidic control of gas turbine injectors. n
• Akash564
Subscriber
Thanks Rob! I meant the distance between the burner and the combustion chamber. I will look into that work.n