RQL modelling: two reactos network and equivalence ratio variation
Alex Svezhov
Hello there,
I am trying to model the RQL (rich-burn, quick-quench, lean-burn) combustor for burning methane-air mixture.
The RQL scheme assumes two regions: rich burn and lean burn. So in Cantera, I am implementing these zone using the reactors network: combustor1 and combustor2 (please see code snippet). The reactors are connected through the valve. The constant total airflow rate (from the reservoir) is also divided into two: by varying the amount of air to the first (rich) combustor, we can achieve different equivalence ratios. The fuel reservoir supplies only the first reactor at a constant rate. The simulation aims to demonstrate how the final temperature and product composition in the second (lean) combustor vary with the equivalence ratios in the fuel-rich region (first combustor).
I am using Cantera 2.4.0 with Python 3.7 on Windows 10. Please see further details in the code/comments.
Issues/questions:
- I noticed that different dt (time increment for the "while-loop", where the reactor network is advanced; line 129) would produce very different results - temperature and composition plots. Is it a problem with the Cantera integrator, or is something incorrect with my implementation?
- I am not sure if I should include the initialization of reactor, reservoirs, and flow controllers within the "for-loop" where I change the equivalence ratio in the rich combustor? Would this inclusion create a new flow device between the system's components at each iteration?
- There is no function advance_to_steady_state for my version of Matlab-Cantera. Is my implementation of reaching the reactor steady-state correct? Are there better alternatives?
Thanks in advance, and I will appreciate any help.