Simulation Models for High Temperature Air Combustion - High Temperature Air Combustion: From Energy Conservation to Pollution Reduction - page 179

adopted in the generic software available in the market. The mixing-is-reacted model,
for example, is suitable for predicting what will happen after combustion has taken
place and not for judging whether or not combustion will take place. Accordingly,
the model is incapable of calculating whether or where combustion takes place in
the furnace. The combustion model to be used for HiTAC simulation must be a
model capable of expressing precise reaction rates in a low oxygen concentration
atmosphere. The dimensions of the reaction zone tend to be thickened in low oxygen
combustion, which is quite different from what common flame sheet models can
predict. Therefore, the one-step global reaction model is not suitable for this new
type of combustion.
Since, as mentioned above, there may be cool material to be heated in a real
furnace, it is important for numerical simulation of practical furnaces to estimate
the heat transfer to the material to be heated as precisely as possible. In a thermal
field where heat loss through the wall or heat transfer to the material to be heated
is involved, the system is no longer taken to be adiabatic. A theoretically complete
combustion temperature based on the adiabatic assumption cannot be applied to the
temperature of burned gas actually recirculating inside the furnace.

3.2.2 P ROPOSED I MPROVEMENTS

As mentioned above, since HiTAC is a new combustion method, its characteristics
cannot be expressed using the combustion models incorporated in the existing
generic software for thermofluid calculation. The following are improvements
required for making software capable of the numerical simulation of HiTAC.

1. Simulation results have to reflect temperature defects caused by thermal

dissociation in the high temperature range above 2000 K, when the full
reaction mechanism is not adopted into the combustion reaction model.
We can consider the following improvements. One is the introduction of
large virtual specific heat into one-step global reaction models to yield
the temperature drop resulting from the actual thermal dissociation in the
high temperature range. Another way to correct the calculated high tem-
peratures caused by the assumption of perfect combustion is to place a
limiting value inferred by the local chemical equilibrium.

2. The rate constants of a reaction model, such as one-step global reaction

models incorporated in the existing generic software for thermofluid sim-
ulations, are not always appropriate for directly applying to HiTAC, since
those constants are mostly adjusted for ordinary combustion using ambient
temperature air. It is incorrect to think there will be no problem, even
when we use a full reaction mechanism, since the accuracy of all the
associated rate constants has not been confirmed, even for the elementary
reactions. Uncertainty about the reaction rate constants increases further
when it comes to composing global reactions or reduced mechanisms.
This is because the reaction rate constants are solely empirical. Therefore,
the constants have to be optimized on the assumption that air temperature
and oxygen concentration are changeable.