COMBUSTION SIMULATION OF A COMBUSTION AND ANALYSIS OF MASS FRACTION OF EACH SPECIES

COMBUSTION SIMULATION OF A COMBUSTION AND ANALYSIS OF MASS FRACTION OF EACH SPECIES

Abstract: The aim of the project is to simulate the combustion of air and fuel mixture. The combustion model used was based on turbulent parameters that depends on turbulent parameters. The fuel mixture used was air-methane mixture. This simulation makes a main assumption that the parameters are axis-symmetric about the geometric axis of the combustor, reducing the simulation to 2-d axis symmetric model. The inlet and outlet of the model are chosen accordingly on the boundaries and the simulation is made to run until convergence is attained.

The types of combustion simulation in fluent are:

Non-premixed combustion: This combustion model is used when both the fuel and air is not mixed before they enter the combustion chamber. This combustion model is used to simulate Diesel Engine and pulverized coal burners.

Premixed combustion: This combustion model, unlike the Non-premixed model, fuel and oxidizer is completely mixed up to the molecular level. This model is used to simulate homogenous charge combustion such as Petrol engine.

Partially pre mixed combustion: This combustion model is an intermediary representation of premixed and non-premixed combustion model.

Species transport: species transport equation is suitable to determine local mass fractions of the involved species in a combustion model. Apart from that it also predicts the rate of production and transfer of particular species.

Composition PDF Transport theory: The composition PDF transport theory is based on a transport equation for the domain. It uses a probability density function to determine the state of any particular species involved such as temperature and RMS velocity in the combustion reaction at any given location in turbulent a turbulent model.

Extracting the solid model:

The 2-d axis symmetric plane is extracted from the 3d model using space-claim modeler using slicing option and the outer contour is obtained. The obtained closed curve is filled and the 2-d surface is exported to the meshing module.

Meshing: The meshing of the entire domain was done using 2-d triangles. Near the inlet, proximity capture is used so as to improve the resolution of the mesh at that particular location only.

The mesh statistics are as follows.

No of nodes : 54884

No. of elements : 108219

Named selections: in the 2-d model, named selections are used to identify the inlet, outlet and axis of the axis-symmetric model.

Fig 1: The mesh model of the 2d axis symmetric model.

Fig 2: The Mesh quality of the computational domain.

Simulation:

After importing the mathematical meshed model into the fluent module, the model is checked for errors. 2d- Axis symmetric option is enabled.

Since the combustion is based on turbulence parameters, k-e turbulence model is used. The turbulence length scale will be based on the ratio of the parameters k and e.

The combustion model chosen is “species transport”. This equation is suitable to determine local mass fractions of the involved species in a combustion model. Apart from that it also predicts the rate of production and transfer of particular species. In case of turbulent flow with diffusion and convection criteria, the local Schmidt number is used in the Governing differential equation. Since only few species are involved, the combustion process is limited to the volume and not in the walls.

Both inlet convection and diffusion are found in this case. The turbulence chemistry interaction relates the turbulence parameters to the combustion criteria. If the criteria is fulfilled, then combustion reaction is predicted by the solver. In our case, Eddy-dissipation model is preferred because the flow is turbulent and the chemical reaction rate is faster compared to transport rate. This model is only used when the flow is predominately turbulent in nature. Another main reason is, since there is a steady supply of air and fuel, the combustion process is self-sustained and steady in the control volume, in other words a stable flame is established.

Boundary conditions:

Air inlet : air at 300 K with normal velocity 0.5 m/s with O2 mass fraction 0.23.

Fuel inlet : pure methane gas at 300 K with inlet normal velocity 80 m/s.

Outlet : Pressure outlet condition with Gauge pressure set to 0 pa.

Reports and convergence:

The simulation is initialized using hybrid initialization and the simulation is carried out until convergence. Temperature contour is updated simultaneously to check for convergence alternatively. The system may be assumed to have attained convergence if the temperature contour does not change after certain number of iterations.

Fig 3: The temperature profile after attaining steady state conditions.

Post processing and results:

In the ANSYS post-processor, the converged temperature and mass fraction contour are checked and validated. Vertical lines covering the span of the domain is created with equal spacing of .2 meters starting from the inlet to the outlet.

NOx : the constituent NOx is used to describe the oxides of nitrogen. These oxides of nitrogen are more likely to form in temperatures greater than 1500 degrees Celsius in the presence of excess oxygen. In this combustion simulation, although the favorable temperature is attained NOx formation cannot be found since NOx model is not used. The no. of constituents were set 5 by default which only includes CH4, CO2, H2O, N2 and O2.

The Line-averaged value of mass fraction is calculated for each individual species and plotted on graph. This line-averaged value can be applied to the entire cylindrical 3-d domain since the problem is axis-symmetric. The line averaged values of the mass fraction of the species at regular intervals are as follows.

Fig 4: Vertical lines along the length of the domain placed at equal intervals in order to determine line averaged value of Mass fraction of each species.

The vertical lines are named from 1 to 11 starting from the origin and placed at an equal interval of .2 metres.

Table 1: Tabulation of line averaged mass fraction at different locations relative to the origin(inlet)

S.no	Distance from origin (metres)	CH4	CO2	H2O	N2	O2
1	0	0.0555556	1.31E-05	1.07E-05	0.727219	0.217202
2	0.2	0.0573892	0.057892	0.0474216	0.709532	0.127765
3	0.4	0.0480683	0.0842801	0.0690304	0.70929	0.0893314
4	0.6	0.0758852	0.138044	0.113058	0.6728	0.000212294
5	0.8	0.105932	0.133752	0.109552	0.650763	4.28E-07
6	1	0.103062	0.134177	0.109898	0.652863	2.83E-08
7	1.2	0.103399	0.134125	0.109855	0.652622	4.67E-09
8	1.4	0.0996133	0.134684	0.110311	0.655391	9.92E-10
9	1.6	0.0973435	0.135019	0.110585	0.657053	2.62E-10
10	1.8	0.0933238	0.135613	0.11107	0.659994	5.29E-11
11	2	0.094773	0.135404	0.110883	0.65894	1.14E-11

The mass fraction contours of the species are as follows.

Fig 5: Mass fraction contour of CH4

Fig 6: Mass fraction contour of CO2

Fig 7: Mass fraction contour of H2O

Fig 8: Mass fraction contour of N2

Fig 9: Mass fraction contour of O2

The variation of Line averaged mass fractions of the species along the domian are as follows.

Fig 10: Line averaged variation of mass fraction of CH4

Fig 11: Line averaged variation of mass fraction of CO2

Fig 12: Line averaged variation of mass fraction of H2O

Fig 13: Line averaged variation of mass fraction of N2

Fig 14: Line averaged variation of mass fraction of O2

Implementation of NOx model for determining mass fraction of NO

For determining the mass fraction of pollutant NO. Thermal NOx model is enabled and the simulation carried out again until convergence is attained. This simulation calculates the mass fraction on NO in the mixture by using Thermal Equillibrium method. It must be noted that the favourable conditions for NOx formation is found at some parts of the domain. The maximum temperature being around 2400 K is quite suitabe for reaction mechanism to produce NO. The simulation is limitted to only formation of NO by excess thermal energy only and other methods that results in the formation of NOx are assumed to be trivial.

The mechanism of formation of thermal NOx according to Zel\'Dovich mechanism.

N₂+ O ⇌ NO + N

N + O₂ ⇌ NO + O

N + OH ⇌ NO + H

In the simulation reduction mechanism of NOx is not considered.

The mass fractions pollutant NO at different locations are as follows.

Table 2: Tabulation of line averaged mass fraction of NO

S.no	Distance from origin	NO
1	0	2.29E-06
2	0.2	0.000651382
3	0.4	0.0006943
4	0.6	0.000829402
5	0.8	0.000662886
6	1	0.000671893
7	1.2	0.000668099
8	1.4	0.000677898
9	1.6	0.000683827
10	1.8	0.000697787
11	2	0.000713327

Fig 15: Mass fraction contour of NO

The line averaged mass fraction of NO at different locations are plotted using CFD post processor.

Fig 16: Line averaged mass fraction of NO