Mixing Tee Simulation

AIM

The aim of this project is to simulate the flow inside a mixing tee joint

OBJECTIVES

The objectives of this project is to simulate flow and study the mixing effectiveness

MIXING TEE JOINT

Mixing tee are flow control devices in which they have two inlets, one for the hot fluid and the other one is for the cold fluid. It is also known as T joint which is used in the mixing process and is used in various applications such as the air conditioner.

The mixing tee simulations were run using the 4 various cases to study the effectiveness of mixing process. The general stages involved in the simulation are

• Geometry
• Meshing
• Setup
• Solution
• Results

FORMULAS USED

Momentum ratio(MR) = velocity at cold inlet / velocity at hot inlet

CASE 1(short mixing tee with MR=2)

• Velocity of hot air = 3m/s
• Velocity of cold air = 6m/s
• Temperature of the hot air = 36C
• Temperature of the hot air = 19C
• momentum ratio = 2

Geometry

The geometry is first created in a CAD software in which the 3D model shows the two inlets one is labelled as inlet X and the other one is labelled as inlet Y. The mixing tee model also has an outlet from which the two mixed fluids flows out of the joint.

The next step in the geometry stage is to extract the wet volume from the original geometry. Wet volume is defined as the volume of the solid model where there is only fluid flow within this volume. In spaceclaim we have to extract the fluid volume by using volume extract command. After extracting the volume we have to supress the solid model and only interact with the fluid volume and so the solid model is supressed.

Meshing

In the meshing stage we have to create the mesh in order to divide the entire geometry into small elements, so the default mesh size of 0.01 m. We are discretising the whole model within which we define as the computational mesh. After this step we have to define the boundaries in order to differentiate the flows coming from the hot and the cold regions.

The number of elements = 15273

Setup

Now we have to set the physics correctly in order to run the simulations, The following setup steps were made

1. Solver - pressure based solver
2. Turbulence model - K-epsilon realizable model
3. Boundary conditions -

• Inlet X velocity = 3m/s
• inlet Y velocity = 6m/s
• Hot air Temperature = 36C
• cold air temperature = 19C

Now we can start the hybrid solver and initialize the solution, then we specify the number of iterations and the solution is started and the fluent solver calculates the flow field parameters

Residual plot shows us the parameters which is being calculated by the solver. Now the residual plot shows us that the solution has run for 300 iterations and we can see that the solution has converged 200 iterations.

Now the standard deviation plot shows us the effectiveness of mixing. A low standard deviation means the mixing process is good, in this case we can see that the standard deviation is around 1.7C which is low and it indicates the mixing process at the outlet is quite good.

Results

Now in order to create the contour plots we define the XY plane and choose the variable for which the plane should show.

Temperature contour

Here we see that the temperature distribution along the length of the pipe is shown where we can see the mixing process and how the temperature is being distributed in the domain.

Velocity contour

the velocity contour also shows us the formation of the boundary layers during the mixing stage and as the velocity decreases along the length of pipe, there is low velocity profiles near the boundary walls.

Temperature along the length of pipe

Temperature across the pipe

Velocity along the length of pipe

Velocity across the pipe

CASE 2(short mixing tee with MR=4)

• Velocity of hot air = 3m/s
• Velocity of cold air = 12m/s
• Temperature of the hot air = 36C
• Temperature of the hot air = 19C
• momentum ratio = 4

Geometry

The geometry is first created in a CAD software in which the 3D model shows the two inlets one is labelled as inlet X and the other one is labelled as inlet Y. The mixing tee model also has an outlet from which the two mixed fluids flows out of the joint.

The next step in the geometry stage is to extract the wet volume from the original geometry. Wet volume is defined as the volume of the solid model where there is only fluid flow within this volume. In spaceclaim we have to extract the fluid volume by using volume extract command. After extracting the volume we have to supress the solid model and only interact with the fluid volume and so the solid model is supressed.

Meshing

In the meshing stage we have to create the mesh in order to divide the entire geometry into small elements, so the default mesh size of 0.005 m. We are discretising the whole model within which we define as the computational mesh. After this step we have to define the boundaries in order to differentiate the flows coming from the hot and the cold regions.

The number of elements = 29575

Setup

Now we have to set the physics correctly in order to run the simulations, The following setup steps were made

1. Solver - pressure based solver
2. Turbulence model - K-epsilon realizable model
3. Boundary conditions -

• Inlet X velocity = 3m/s
• inlet Y velocity = 12m/s
• Hot air Temperature = 36C
• cold air temperature = 19C

Now we can start the hybrid solver and initialize the solution, then we specify the number of iterations and the solution is started and the fluent solver calculates the flow field parameters

Residual plot shows us the parameters which is being calculated by the solver. Now the residual plot shows us that the solution has run for 300 iterations and we can see that the solution has converged 175 iterations.

Now the standard deviation plot shows us the effectiveness of mixing. A low standard deviation means the mixing process is good, in this case we can see that the standard deviation is around 1.3C which is low and it indicates the mixing process at the outlet is quite good.

Results

Now in order to create the contour plots we define the XY plane and choose the variable for which the plane should show.

Temperature contour

Here we see that the temperature distribution along the length of the pipe is shown where we can see the mixing process and how the temperature is being distributed in the domains.

Velocity contour

the velocity contour also shows us the formation of the boundary layers during the mixing stage and as the velocity decreases along the length of pipe, there is low velocity profiles near the boundary walls.

Temperature along the length of pipe

Temperature across the pipe

Velocity along the length of pipe

Velocity across the pipe

CASE 3(long mixing tee with MR=2)

• Velocity of hot air = 3m/s
• Velocity of cold air = 6m/s
• Temperature of the hot air = 36C
• Temperature of the hot air = 19C
• momentum ratio = 2

Geometry

The geometry is first created in a CAD software in which the 3D model shows the two inlets one is labelled as inlet X and the other one is labelled as inlet Y. The mixing tee model also has an outlet from which the two mixed fluids flows out of the joint.

The next step in the geometry stage is to extract the wet volume from the original geometry. Wet volume is defined as the volume of the solid model where there is only fluid flow within this volume. In spaceclaim we have to extract the fluid volume by using volume extract command. After extracting the volume we have to supress the solid model and only interact with the fluid volume and so the solid model is supressed.

Meshing

In the meshing stage we have to create the mesh in order to divide the entire geometry into small elements, so the default mesh size of 0.01 m. We are discreatizing the whole model within which we define as the computational mesh. After this step we have to define the boundaries in order to differentiate the flows coming from the hot and the cold regions.

The number of elements = 29575

Setup

Now we have to set the physics correctly in order to run the simulations, The following setup steps were made

1. Solver - pressure based solver
2. Turbulence model - K-epsilon realizable model
3. Boundary conditions -

• Inlet X velocity = 3m/s
• inlet Y velocity = 6m/s
• Hot air Temperature = 36C
• cold air temperature = 19C

Now we can start the hybrid solver and initialize the solution, then we specify the number of iterations and the solution is started and the fluent solver calculates the flow field parameters

Residual plot shows us the parameters which is being calculated by the solver. Now the residual plot shows us that the solution has run for 300 iterations and we can see that the solution has converged 250 iterations.

Now the standard deviation plot shows us the effectiveness of mixing. A low standard deviation means the mixing process is good, in this case we can see that the standard deviation is around 0.25C which is low and it indicates the mixing process at the outlet is quite good.

Results

Now in order to create the contour plots we define the XY plane and choose the variable for which the plane should show.

Temperature contour

Here we see that the temperature distribution along the length of the pipe is shown where we can see the mixing process and how the temperature is being distributed in the domain.

Velocity Contour

the velocity contour also shows us the formation of the boundary layers during the mixing stage and as the velocity decreases along the length of pipe, there is low velocity profiles near the boundary walls.

Temperature along the length of pipe

Temperature across the pipe

Velocity along the length of pipe

Velocity across the pipe

CASE 4(long mixing tee with MR=4)

• Velocity of hot air = 3m/s
• Velocity of cold air = 12m/s
• Temperature of the hot air = 36C
• Temperature of the hot air = 19C
• momentum ratio = 4

Geometry

The geometry is first created in a CAD software in which the 3D model shows the two inlets one is labelled as inlet X and the other one is labelled as inlet Y. The mixing tee model also has an outlet from which the two mixed fluids flows out of the joint.

The next step in the geometry stage is to extract the wet volume from the original geometry. Wet volume is defined as the volume of the solid model where there is only fluid flow within this volume. In spaceclaim we have to extract the fluid volume by using volume extract command. After extracting the volume we have to supress the solid model and only interact with the fluid volume and so the solid model is supressed.

Meshing

In the meshing stage we have to create the mesh in order to divide the entire geometry into small elements, so the default mesh size of 0.01 m. We are discreatizing the whole model within which we define as the computational mesh. After this step we have to define the boundaries in order to differentiate the flows coming from the hot and the cold regions.

The number of elements = 29575

Setup

Now we have to set the physics correctly in order to run the simulations, The following setup steps were made

1. Solver - pressure based solver
2. Turbulence model - K-epsilon realizable model
3. Boundary conditions -

• Inlet X velocity = 3m/s
• inlet Y velocity = 12m/s
• Hot air Temperature = 36C
• cold air temperature = 19C

Now we can start the hybrid solver and initialize the solution, then we specify the number of iterations and the solution is started and the fluent solver calculates the flow field parameters

Residual plot shows us the parameters which is being calculated by the solver. Now the residual plot shows us that the solution has run for 140 iterations and we can see that the solution has converged 140 iterations.

Now the standard deviation plot shows us the effectiveness of mixing. A low standard deviation means the mixing process is good, in this case we can see that the standard deviation is around 0.25C which is low and it indicates the mixing process at the outlet is quite good.

Results

Now in order to create the contour plots we define the XY plane and choose the variable for which the plane should show.

Temperature contour

Here we see that the temperature distribution along the length of the pipe is shown where we can see the mixing process and how the temperature is being distributed in the domain.

Velocity contour

the velocity contour also shows us the formation of the boundary layers during the mixing stage and as the velocity decreases along the length of pipe, there is low velocity profiles near the boundary walls.

Temperature along the length of pipe

Temperature across the pipe

Velocity along the length of pipe

Velocity across the pipe

MESH INDEPENDENT STUDY

Now for the first two cases in the short mixing tee simulations the mesh size is varied for the second case to 0.005m and for the first case the mesh size is taken as 0.01m. Now we see that in both the cases the solution is converging after a set number of iterations. So thereby we can conclude that the solution is mesh independent. Mesh independence means the solution should converge irrespective of the mesh size or adding of the inflation layers. The mesh inflation is added in the second case with momentum ratio as 4. In this solution in the residual plot we see that the solution has converged for a steady state value.

COMPARIOSION

parameter	short mixing tee with MR=2	short mixing tee with MR=4	long mixing tee with MR=2	long mixing tee with MR=4
cell count	15273	15273	29575	29575
average outlet temperature(degree celsius)	30.25	31.5	27.75	27.75
number of iterations for convergence	200	175	250	140

CONCLUSION

1. In the short mixing tee simulations when we increase the momentum ratio from 2 to 4 we see the mixing process is more efficient as there is also drop in the standard deviation
2. Compared to long and short mixing tee simulations the long mixing tee should have better mixing than the short pipe.
3. When the velocity in the pipe is increased from 6m/s to 12 m/s for the cold air the mixing efficiency increases. So we can conclude that the mixing efficiency is directly proportional to momentum ratio increase
4. Also, the number of iterations required for convergence decreases with increase in momentum ratios.
5. As the momentum ratio increases, we see a drop in the standard deviation, so the mixing efficiency increases.