Dear All,

I have a premixed combustion simulation i am running. I have done the cold flow and i want to switch to the hot flow. I need to patch only the combustor and so to do this, i need to put some products (H2O and CO2) into the domain inside the combustor since i am using the eddy dissipation model. How can i do this in CFX? I am thinking of using "Expression" but when i type the following Temp=1800[K]*step((x-0.8)/1[m])+300[K]*step((0.8-x)/1[m], it is not running. My combustor axis is along the x axis and my combustor length is 0.8m and my x=0 point is at the entry to the combustor.

When i run the simulation, i keep getting the following error messages:

"Error processing expression 'Temperature'. Inconsistent dimensions on each side of "-" operator at position 16.

Dimension on left: 'm' Dimension on right: '<Dimensionless>' Error processing expression: Temperature = Temp.

Any help on how to do this?

Thanks.

Jimmy

I do not have much knowledge of combustion but, as far as the your expression it need to be slightly modified as 1800[K]*step((x-0.8[m])/1[m])+300[K]*step((0.8[m]-x)/1[m]

you need to have same dimensions on the both side of any operator.

Hi and thanks for the help. I made the modification and it seems to work. but i am not clear about how this step function works in CFX. Maybe i can explain what i want to do so you can see whether i did the right thing in the expression. My geometry is such that i have the origin at the entry to my combustion chamber and so my combustor is in the positive x direction and mixing chamber with the fuel and air inlets in the negative x direction. The combustor length is 0.8m. So what i want is set the temperature everywhere inside the combustion to 1800K and to set the temperature in the mixing zone to 300K. Any idea whether what i have is right?

The reason i have to do it this is because i only want burning inside the combustor and so if i leave the initialization temperature to "automatic", there will be burning in my entire domain.

Any suggestion will be appreciated.

Thanks.

Jimmy

Dear Jimmy, the problem with your equation is for the value of x= 0 -> 0.8, the first step function always gives a zero value

i have modified your equation little bit 1500[K]*(step((0.8001[m]-x)/1[m])*step(x))+300[K]

you can try this and let me know if it works, but there is a slight problem that at x=0 the output will be 1050K, but for the rest of the domain i hope it works as per requiremnt

you can just think a bit more and can avoid the problem at x=0