[tellmetruth]

Dear all,
I am a fresh user of fluent and doing some simulation to study the surface influent in the microflame with it.

In the 2-D planar channel, I implement the gas-phase chemical kinetics is based on GRI-Mech 3.0 excluding Ar, C3H8 and C3H7 species and associated reactions (50 species, 309 reactions). The surface reaction mechanism contain 10 reactions and 5 species proposed by Raimondeau.
For BCs, velocity inlet: air ratio = 0.95 ,v is constant and uniform 0.97m/s and T = Tw
asymmetry axis, non slip and neglect thickness of wall, Tw=constant.
outlet: outflow( the BC for outlet maybe a source of problem)
SIMPLE Pressure-velocity coupling, remain default in other setting.
UNSTEADY SOLUTION, time step is 1e-5s

After hundreds of times' simulation, all the results related to the species distributions (including the species in the gas phase and surface sites) are NOT monotonic and unreasonable.

Even worse

BTW, the X axis is along the planar channel.
I know my description is not fully detailed. BUT, the point is the reference specie distribution should be smooth and single peak or monotonic. I just want to know how I should modify the settings to remove the rough and steep on those species distribution curves and get the reasonable ones.
The reference curves from other papers( hope its legal to put them here)


If someone can give me some tips, you would really help me.
LIU
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Sorry, because I don't know whether those pictures are visible, I share the pictures as links to my cloud driver.
https://drive.google.com/open?id=0By...XJqdVlzMEp6bVUhttps://drive.google.com/open?id=0By...XJqdVlzMEp6bVU
https://drive.google.com/open?id=0By...UxkeHFDaTZiSm8
https://drive.google.com/open?id=0By...EloYm5kOWsyUFE
https://drive.google.com/open?id=0By...nJra3RqTlI1M28