Hi everybody. I have one big problem. I cant reach convergence neither in steady state nor in transient. RMS are about 10^-3, so very poor. The other problem is temperature. From literature I know that for such cases (pulverized lignite is burnt) max temp should be about 1700-1800C. In my case I have something about 2200-2700C. What am I doing wrong? Please about Your help.

Hi,

Accurate combustion modelling is tricky, you will need to validate the simulation one step at a time to get accurate answers.

Glenn Horrocks

Hello,

there are many parameters that affect convergence. What are your coal and air mass flow rates, particle sizes (Rosin-Rammler parameters) and number of particles? Do you use default CFX values for devolatilization and char combustion? From my experience, you could get quite good results with gross particles using CFX defaults. May be that is the point you should start from.

Regards

I am using default settings of HC fuel. I have changed only radical parameters (proximate and ultimate analysis) and heat combustion. In this furnace there is many inlets of air and air-pulverized lignite mixture- about 20 at all. Total power is about 1 GW - power released from lignite burning. This furnace is in use in power station in Poland. Rosin Ramler Power: 1.1 and size: 0.0001 m. Maybe the problem is in my grid-I have tetra mesh with about 1 million elements. There are big differences between sizes of this elements-about 10 times. It was only way to cover properly relatively small nozzles in comparison with dimensions of furnace (50 m hight). I even started from pure flow without reaction and obtained good convergence (in transient mode) but with combustion still I have no convergence and so high temperatures. Do You have maybe some other ideas. Please about help.

How many particles do you track?

We use hexa. You may try to change Arrhenius coefficients, maximum flame and particle temperature and use temperature damping. In case you have not tried yet.

I was thinking about hexa but problem for me is big diferences in dimensions. With hexa probably I would have to generate very fine mesh. Temperature damping is usefull for finite rate chemistry - I just thing follow ansys manual. I use eddy dissipation combustion model. I have tried maximum flame temperature but it is working very strange. I set max temp 2000K and still observing temperatures about 2200K. On the other hand it is artifficial solution. I though that the simmulation would tell me something about max temperatures not me. ;-) Thanks for answer but I am still searching something different. Maybe I am wrong in something.

What number of particles is the most sufficient? I track about 4800 particles. I my furnace I have 24 inlets of coal-air mixture, so it is 200 particles per one inlet. Mass flow of coal is 3.48 kg/s per one inlet. I dont have any idea if it is good number. May it have something to do with convergence?

Hi Jarek,

pl specify ur radiation model and boundary conditions to the wall, abs coeffi, scatt coeff,

radiation model P1 but I have tried monte carlo and discrete transfer too. boundary condition on the walls is constant temperature 700K. What are the abs and scatt coefficients?

it is absorption coeff and scattering coefficient for gas phase radiation.