Mixing efficiency using Ansys Fluent.
Three cases are presented here one by one and comparative observations are written in the end. For all the cases the number of mesh elements is kept almost the same and hot and cold temperature is taken as 25 0C and 10 0C resp. For all the cases hot inlet is maintained in x-direction and cold inlet in a y-direction.
CASE I:
Short pipe with a hot inlet at 3 m/sec and momentum ratio 2, which gives the cold inlet velocity 6 m/sec.
The fluid domain is extracted from the given model geometry and meshing is done with Ansys meshing by keeping element size as 0.0012 m. The reason for keeping the size as minimum as possible is to get the good temperature and velocity profile at the outlet.
The 5 inflation layers are added at the wall boundary to see if there is any change at the outlet and to capture the details at the wall. The mesh with the number of elements 458247 looks like fine mesh to proceed.
The simulation ran for 450 iterations for a safe side but, the plots of outlet temperature and standard deviation in outlet temperature shows that 350 iterations have been more than enough. It could have saved simulation time. In fact, I ran the simulation for 350 iterations and difference in outlet temperature is almost negligible. Plots are shown below:
Residual Plot:
Average outlet temperature plot.
The standard deviation in the outlet temperature plot.
The standard deviation seems to be around 1 0C at the outlet and going below that would be either time consuming or not even possible. I ran one simulation for 800 iterations but, it still around the same. The 450 iteration of the above case gives the following velocity and temperature profiles at the XY plane in the middle of the pipe.
Temperature Profile:
Velocity Profile:
The average outlet temperature in the above case is coming around 20.084 0C. The mesh independence study has been done from varying the cell count from 1 lakh to 4.5 lakh and temperature at the outlet is not varying much but, profile looks very smooth with more number of elements.
CASE II:
A long pipe with a hot inlet at 3 m/sec and momentum ratio 2, which gives the cold inlet velocity 6 m/sec.
The overall process done for this case is the same as above. The element size kept at 0.00137, which gave the cell count 458945. The inflation layer added at the wall which deteriorating the quality of the mesh but, it does not seem to have any effect on the outlet temperature.
The simulation ran for 350 iterations and plots for the same are:
Residual Plot:
Average outlet temperature plot:
The standard deviation in the outlet temperature plot:
The simulation gave the following outlet profiles.
Temperature Profile:
Velocity Profile:
The average outlet temperature in the above case is coming around 20.111 0C. The long pipe seems to have no impact on the average outlet temperature So, we can safely conclude that extending the length is unnecessary.
CASE III:
A long pipe with a hot inlet at 3 m/sec and momentum ratio 4, which gives the cold inlet velocity 12 m/sec.
The overall process done for this case is also the same as above. The element size kept at 0.0012, which gave the cell count 458681. The simulation ran for 450 iterations as the velocity from the cold inlet is higher which has created more turbulence in the flow. The plots are as follows.
Residual Plot:
Average outlet temperature plot:
The standard deviation in the outlet temperature plot.
The simulation gave the following outlet profiles.
Temperature Profile:
Velocity Profile:
As legends to be maintained same for all three cases, the third case velocity is higher that legend used and red color indicated its higher than 9m/sec. The average outlet temperature in the above case is coming around 17.489 0C. The higher velocity is responsible for inducing more turbulence in the flow which mixing the flow properly and the temperature is dropping by 3 0C than the previous two cases.
Conclusion:
|
Sr. No. |
Cell Count |
Average Outlet Temp.(0C) |
Number of Iteration |
CASE Number |
|
1 |
458247 |
20.084 |
450 |
1 |
|
2 |
458945 |
20.111 |
350 |
2 |
|
3 |
458681 |
17.489 |
450 |
3 |
The grid independence study has been done and 1 lakh cell count for all the three cases seems to be a good choice as far as the average outlet temperature is concerned. But, cell count more than 4 lakhs gives good profiles. I tried to keep cell count almost same for all the three cases so that comparison would be easy. Taking larger length for pipe does not make any difference in outlet temperature as well as velocity and temperature profile. Increasing the velocity at cold inlet helps increase mixing efficiency and temperature at the outlet seems to be less than the first two cases. Inflation layers are added to see if it makes any difference in the overall result but it does not. Because we need not capture anything at the walls. So inflation does not have any role in the overall mixing and temperature outlet.