combustion modelin with SORSCA
 

Hi everybody. I am writing a Weller-based combustion model by mean of sorsca.f. I have defined the various scalars in Prostar with relative heat of formation, and activated the chemico-thermal enthalpy. I am facing the following problem: if I define the source terms for every scalar except for the products (CO2 and H20), the latter being imposed on the value of the other scalars by changing tg(i,is-1), the results seems quite reasonable, but the info file gives me the warning "scalar equation "i" not converged" , where "i" refers to CO2 and H2O, and towards the end of the simulation the combustion "freezes".

If I define a sorce term for CO2 and H2O, nothing happens, except for an unrealistic temperature drop in the initial burnt zone. I have checked the source terms, and the total reactants souce is equal to the total products source, except for the sign.

Could anybody help me to understand what is happening? Any help really apreciated. Roadracer

Re: combustion modelin with SORSCA
 

It's not clear how you are setting the values of the CO2 and H2O scalars. Are you writing directly to the T array (very bad practice) or using SCALFN (good practice)?

Re: combustion modelin with SORSCA
 

I have tried both. Defining the scalars CO2 and H2O changing directly T(ipstar,is+1) leads to reasonable results. Defining the scalars CO2 and H2O by mean of scalfn (i.e. defing s1p) leads to strange results. In the latter case the sum of s1p for the various reactants is equal to the sum of s1p for the products. s2p is always 0 in my calculation.

Re: combustion modelin with SORSCA
 

My suggestion was to use SCALFN for the "algebraic" scalars. I think you read it as SORSCA, as you talk of setting S1P. SORSCA is for source terms, whereas SCALFN explicitly sets the scalar value by returning it in PHI. That's a better way to set an algebraic scalar through user coding.

Re: combustion modelin with SORSCA
 

Hi thank you for the suggestions, but I found the error. I didn´t save a variable. Now it works fine (at least fo homogeneous fuel distribution) using sorsca. Cheers Marco