Near the beginning of this month we made a significant leap in progress with a change in our ignition method. We replaced the spark ignition method, which had caused us so much trouble, with a patched temperature on a few cells at the ignition point. This resulted in significantly more reliable results and we produced some promising simulations.
Unfortunately, we never got to the bottom of why the spark ignition method was unsuccessful. Project SUSANA documentation simply stated that more research was required on the method and this appears to still be the case.
The rest of the month was spent further refining this model. Amongst other smaller changes, we swapped to a more detailed Chemical Kinetic model. This meant that more chemical reactions were simulated on top of the simple Hydrogen + Oxygen = Water which we had been using. This came with its own issues, namely the lack of any reaction after ignition, but worked well after some adjustment and it yielded more realistic temperature values.
We also created a spreadsheet in order to calculate the flame velocity at regular intervals throughout the duct so that we can compare flame speed to the HYCOM data. Using this as validation we completed a full simulation with a smaller timestep and began a mesh and timestep sensitivity analysis.
~Iain~
Comments