i am trying to simulate the turbulent boundary layer combustion inside the combustion chamber of a hybrid rocket..its a diffusion flame happening at 5 bar chamber pressure..the fuel is 1,3-butadiene and the oxidiser is pure oxygen..as the first step i am having only gases enter the chamber..
for combustion i am using laminar finite rate model.. for turbulence, i had till date been using k-eps RNG model with the default parameters..i tried with laminar model also..
the reaction follows 2 step chemistry..CO forms in the first non-reversible reaction along with H2O and the second reversible reaction is the oxidation of CO..
the activation energy values and the pre-exponential factor values have been taken from a tech report by C L Merkle and S Venkateswaran, Pennsylvania state univ..
i have also been trying with single step model..
with both the models i have not yet been able to get a solution..
i first do the flow alone switching off the reaction..the moment i switch on the reaction the residuals start zooming up..the energy and continuity residuals go above 1E+2 and stays steady..
i am using coupled solver (density based)..with the seggregated solver (pressure based) the reactions get blown off..(that is the reaction rates gradually become low and there are no more reactions after sometime..even with patching the whole zone with a high temp..as high as 3000K..)
initially i worked with courant numbers of the order of 0.1 to 1..later i reduced them to 0.0001 and tried the whole thing..nothing much happened..
what could be the possible problems...???why is that the reactions get blown off in the pressure solver (with patching) where as they are not so with the density solver (even with out patching)..???which turbulence model is prefered for the boundary layer combustion..???pdf models are accurate, but this laminar finite rate model shud predict the flame, atleast at a wrong value, right..??
any suggestions..??or references..??
thanks a lot..
|All times are GMT -4. The time now is 01:29.|