Initial or boundary? there are different. Are you imply that the boundary condition for the flow has changed by time as a function determined from a UDF?

Hi Shirrup. I guess the code you write determines the heat transfer rate as a function of some simulated quantity at each itteration. Lets imagine that the heat flux you evaluate is dependant on the external temperature (I've done similar code to evaluate the heat transfer rate from a human body). The heat transfer rate you calculate with the UDF will push the external temperature in a certain direction, then this temperature will push the heat flux (evaluated by your UDF) in the opposite direction, and so the simulation will diverge. To remedy this problem you need to relax the heat flux before supplying it to the solver. Thats what I did in my case. In essence you change the heat flux not with the whole step predicted, but only with a fraction of it. Lets say Q is the output of your UDF, then:

Q(i+1) = Q(i) + [ Q(i+1) - Q(i) ].r,

i - itteration number, r - relaxation coefficient between 0 and 1, ' = ' is an assignment operator.

Thank you for your replies. Well i will make clear my problem statement. As mentioned in the query the simulation software I am using is FLUENT. The heat transfer equation i am giving as an input as UDF as heat flux as we give in FLUENT. so this UDF is basically an initial input boundary condition. And as per i have read about UDFs for every iteration the heat transfer rate and temperature at the wall boundary should be calculated.

Now how do i calculate the inputs to the equations. All the flow parameters for each equation are taken from the cell itself which FLUENT calculates, like pressure, temperature, density and others. So dis is how the input for the equations are given to.

As u have mentioned George, yes the heat transfer rate is determined as each iteration as the function of the atmospheric properties. But since u said about relaxing the heat flux before input, my doubt is how to do so in FLUENT. Means which option to change. And how to determine the direction of heat flux. Has it to be done first using a pen and paper?

I hope I have made myself clear in my query. Waiting for the replies. Need replies fast as the code I am using it for my dissertation. Thank you.

Try to decrease the time interval see how it works

I am quoting you "Now how do i calculate the inputs to the equations. All the flow parameters for each equation are taken from the cell itself which FLUENT calculates, like pressure, temperature, density and others. So dis is how the input for the equations are given to. "

If i understand you well, This is not a boundary condition

Any thing to be solved for in the course of your solution is not strictly a boundary condition. Yet, if we consider it as a constraint (not BC), you can use the most updated solution(sometimes with harsh under-relaxation) to feed into this constraint to reach convergence