Hi Etienne,
For the ImpingingJet case, this should work fine; you can set the surface temperature to a value of your choice. While the catalytic_combustion.py example shows the case where surface reactions are present, the only change required to model an inert surface is to omit the surface=... argument when constructing the ImpingingJet object. You can also dispense with the more complex solve process used in that example where the reaction rate multipliers are only gradually raised.
I’m not sure what you are referring to in the documentation about adding a surface at a certain location. Both of your ideas for how to represent heat loss could be implemented in the Cantera solver. They would need to be implemented in the C++ core, probably as additions to the Flow1D::evalEnergy method. In the first case (sink term), you’d be adding an additional term to the energy equation at each point. In the second, it sounds like you’d be replacing the energy equation at certain points with a fixed-value condition. The only thing to watch for in that case is that you also set the corresponding entries of the diag vector to indicate that these are now algebraic rather than differential variables in the system. Other than that, most of the work would be around parameterizing these losses and introducing methods for setting those parameters.
Regards,
Ray