Mixing efficiency
# Aim of the project is to simulate flow through mixing tee's of 2 different length for different momentum ratio's and compare the mixing efficiencies.
Mixing tee's are the ones which is commonly used in air conditioning systems. Usually a stream of hot air and stream of cold air (relativley less temperature than hot air) is mixed so as to obtain a desired temperature at the output. Inlets of the stream's are at an angle of 90° and hence it looks like a 'T' junction. By adjusting the flow rate of chilled air desired temperature output can be obtained.
* Setup - two mixing tee's of different length, one short (length = 0.19 meter) and another long (length = 0.27 meter) is selected.
Simulations of air flowing through above mixing tee's are done at momentum ratio of 2, 4.
A typical illustration of mixing tee is as shown below:
Momentum ratio = Velocity at cold inlet / Velocity at hot inlet.
# Steps involved in the simulation as follows:
1. Extracting the fluid volume using ansys spaceclaim.
2. Meshing the geometry in the meshing module of ansys.
3. Setting up the simulation in ansys fluent, which invloves applying boundary condition, selecting appropriate turbulence models, Running the calculations.
4. Post processing the results in CFD post and comparing the result obtained with the help of different contour plots and line plots.
# Mixing tee short:
Fluid volume of short mixing tee when extracted from spaceclaim looks as follow:
After meshing the above mixing tee will look like as follow, mesh element size of 0.008 m is used and also the mesh quality is checked ,
above histogram shows the quality of mesh more than 34% (which is very much higher than the required quality of 5%) hence we can proceed with the above mesh for further simulation.
# comparing the tubulence models:
In here for the mixing tee short, for momentum ratio = 2, simulation is carried using k-ε as well as k-Ω models ,
For momentum ratio of 2, we have v1 = 3 m/s, T1 = 36 °C and v2 = 6 m/s, T2 = 19 °C
For k-ε model: Area weighted average temperature at outlet = 30.25 °C
No.of iterations to converge = 240.
Total No. of mesh/cells = 12756.
For k-Ω model: Area weighted average temperature at outlet = 30.25 °C
No.of iterations to converge = 375.
Total No. of mesh/cells = 12756.
from the above observation it is clear that k-ε took less number of iterations to converge than k-Ω, also in this simulation we are not capturing parameters near wall
which k-Ω does very well, instead we are capturing parameters far from wall which is handeled very well by k-ε, hence all further simulations will be done using k-ε model.
* Mixing tee short - for momentum ratio = 2 and for bondary conditions v1 = 3 m/s, T1 = 36 °C and v2 = 6 m/s, T2 = 19 °C
we have Area weighted average temperature at outlet = 30.25 °C
No.of iterations to converge = 240.
Total No. of mesh/cells = 12756.
# Plots:
* Convergence plot:
* Temperature contour plot along and across the mixing tee:
* Velocity contour plots along and across the mixing tee:
* Line plot of temperature along and across the mixing tee:
* Line plot of velocity along and across the mixing tee:
* Mixing tee short - for momentum ratio = 4 and for bondary conditions v1 = 3 m/s, T1 = 36 °C and v2 = 12 m/s, T2 = 19 °C
we have Area weighted average temperature at outlet = 27.5 °C
No.of iterations to converge = 125.
Total No. of mesh/cells = 12756.
# Plots:
* Convergence plot:
* Temperature contour plot along and across the mixing tee:
* Velocity contour plots along and across the mixing tee:
* Line plot of temperature along and across the mixing tee:
* Line plot of velocity along and across the mixing tee:
# Mixing tee short:
Fluid volume of short mixing tee when extracted from spaceclaim looks as follow:
After meshing the above mixing tee will look like as follow, mesh element size of 0.008 m is used and also the mesh quality is checked ,
from above histogram it can be observed that , mesh quality is more than 32% and hence we can proceed with the generated mesh for further simulation.
* Mixing tee long - for momentum ratio = 2 and for bondary conditions v1 = 3 m/s, T1 = 36 °C and v2 = 6 m/s, T2 = 19 °C
we have Area weighted average temperature at outlet = 30.3 °C
No.of iterations to converge = 230.
Total No. of mesh/cells = 15483.
# Plots:
* Convergence plot:
* Temperature contour plots along and across the mixing tee:
* Velocity contour plots along and across the mixing tee:
* Line plot of temperature along and across the mixing tee:
* Line plot of velocity along and across the mixing tee geometry:
* Mixing tee long - for momentum ratio = 4 and for bondary conditions v1 = 3 m/s, T1 = 36 °C and v2 = 12 m/s, T2 = 19 °C
we have Area weighted average temperature at outlet = 27.5 °C
No.of iterations to converge = 155.
Total No. of mesh/cells = 15483.
# Plots:
* Convergence plot:
* Temperature contour plot along and across the mixing tee:
* Velocity contour plots along and across the mixing tee:
Line plots of temperature along and across the mixing tee:
* Line plots of velocity along and across the mixing tee:
# Observations from the plot:
* From both velocity and temperature plots it can be observed that in mixing tee where the fast moving stream and slow moving stream's meet at that point the velocity as well as temperature decreases drastically. And then it increases gradually towards the outlet.
* The number of cells adjacent to the point of intersection where the value of temperature and velocity is low depends on the momentum ratio.
* Higher the momentum ratio more number of cells adjacent to the intersection will have lesser value of temperature and velocity.
# Grid independence study:
Grid independence study is done on mixing tee short at momentum ratio of 2
1. Size of cells is taken as 0.005 m hence we have total No. of cells = 14228.
for the above number of cells calculations are carried and the average outlet temperature and standard deviation of temperaure is observed. Below are the plots of the same
*Plot of Area weighted average temperature:
* Plot of standard deviation of temperature:
2. Size of cells is taken as 0.003 m hence we have total No. of cells = 39815.
for the above number of cells calculations are carried and the average outlet temperature and standard deviation of temperaure is observed. Below are the plots of the same
* Plot of area weighted average temperature at the outlet:
* Plot of standard deviation of temperature at outlet:
* Conclusion:
From the above plots it is clear that even if the number of cells varied the average output temperature is 30.25 °C and standard devation of temperature at the outlet is around 1.7 °C and hence we can conclude that simulation is grid independent.
# Table of comparison:
| Mixing tee | Momentum ratio | Cell count | No. of iterations to converge | Avg. outlet temperature (°C) |
| Short mixing tee | 2 | 12756 | 240 | 30.25 |
| Short mixing tee | 4 | 12756 | 125 | 27.5 |
| Long mixing tee | 2 | 15483 | 230 | 30.3 |
| Long mixing tee | 4 | 15483 | 155 | 27.5 |
* Conclusion:
Suppose we are desired of having a output temperature of 27 ° C a momentum ratio of 4 in both short mixing tee as well as long mixing tee will give us an output temperature of 27.5 °C
so considering the design aspect that optimizing the length for required output temperature then mixing tee of shorter length will be a good to go option.