CONJUGATE HEAT TRANSFER SIMULATION
ABSTRACT: This report contains the study of a project on the Conjugate Heat transfer Simulation. A case for Conjugate Heat Transfer was setup in Converge with suitable conditions. Input files were simulated using Cygwin and the results were post processed using Paraview and Converge. Study was done to analyse the effect of Grid size and supercycling time duration on the simulation with base grid.
INTRODUCTION: Whenever more than one mode of heat transfer occur, it is called as conjugate heat transfer. In this report a conjugate heat transfer involving a hollow cylindrical pipe of certain thickness was used for the flow analysis. The air in the pipe was flowing from one end and pipe was being heated from outside by a external heat flux. The solid section of the pipe was getting heated up with conduction and the air inside the cylinder was heated at the interface due to convection. The heat was transfered within the air during the flow.
CASE SETUP: Case setup was done in converge. The geometry was made and then mesh was created. after this the geometry was checked for any errors using diagnosis tool. At the begining there were non-manifold edges in the geometry but once the case setup was completed, those errors were ignored by the converge since it realized this was Conjugate Heat Transfer problem. This was followed by assigning meaningful boundaries to the geometry.
Geometry
The next step was to begin the case setup
APPLICATION: General flow
MATERIALS: Solid simulation-Aluminium, Species- solid (Aluminium) and Gas( O2 and N2)
SIMULATION PARAMETERS: Run parameters- Transient, full hydrodynamic.
Simulation Time parameters: End time-0.5 sec, initial and max time step-1e-7 sec
BOUNDARY CONDITIONS: Solid outer wall- Wall, Stationary, slip, heat flux=-10000 W/m^2
Solid side- Wall, stationary, slip, zero normal gradient.
Inlet- Inflow, velocity=3.66 m/s along Z-direction, Species (O2 and N2)
Outlet- Outflow, specified value, Pressure=101325 Pa, species backflow(O2 & N2)
Interface- Interface, forward boundary- fluid, reverse boundary- solid
INITIAL CONDITIONS AND EVENTS: Region ID 0- Fluid, stream id 0
Region ID 1- Solid, stream id 1
PHYSICAL MODEL: Turbulence modeling- Default
Super-cycle modeling- time length for each cycle stage=0.01sec, Conduction CFL number=100
GRID CONTROL: Three grid sizes were used
Case 1- dx=dy=dz=0.004m
Case 2- dx=dy=dz=0.0035m
Case 3- dx=dy=dz=0.003m
OUTPUT/POST PROCESSING: Post variable selection- +Y added to turbulence/combustion
Output files- Generate wall boundary averaged output also checked
USER DEFINED FUNCTION: Default
RESULTS: Following results were obtained on the three different grid sizes on Convergence of temperature, density, mass flow rate, temperature, velocity.
1. Mesh size: Three mesh sizes were compared for the simulation. A slice of the existing geometry was created in Paraview with normal towards Y-axis.
Case 1: dx=dy=dz=0.004m
Case 2: dx=dy=dz=0.0035m
Case 3: dx=dy=dz=0.003m
Idreg is used to indicate the region id. The color blue indicates fluid region and the color red indicates solid region
2. Convergence of temperature:
Case 1:
Case 2:
Case 3:
The above plots show us that the temperature in all three cases converges to a particular value after a certain interval of time. This implies the solid section of the cylinder has achieved its final temperature due to the heat flux. This is the averaged data, therefore max and min temperature will not be same as the mean temperature.
3. TEMPERATURE: The temperature of the solid varies due to the heat flux given and it takes sometime to achieve final temperature. Whereas inside the cylinder the air is heated by convection. Boundary_id_3 shows the temperature at inlet and Boundary_id_4 shows the temperature at outlet. The inlet temperature remains constant fairly over the range but outlet temperature varies and takes some time to gain a stable value. The temperature distribution on the slice is also shown. From the results it is evidently clear that the temperature of fluid is max at the interface and is min at the centre of the cylinder, also the temperature shows a parabolic profile.
Case 1:
Case 2:
Case 3:
4. DENSITY: The density of air vaires with temperature and since this case involves the transfer of heat and mass both the density varies during the flow. There is an initial time lag in change of density at outlet, this is due the time taken by the solid to heat up(since solids have higher specific heats). As the time passes, density attains a constant value. The desnity at outlet is less than the density at inlet. This is simply because of the higher average temperatures at outlet than at inlet.
CASE 1:
CASE 2:
CASE 3:
5. VELOCITY: Velocity at inlet was specfied by us at the time of giving boundary conditions. The velocity was determined with the help of inlet reynold's number(Re=7000). At 300 Kelvin the properties of air density=1.77 kg/m^3, diameter of inlet=0.03m, dynamic viscosity=1.846e5 Kg/m-s. Using equation `Re=rhoVD/mu` we get velocity at inlet to be 3.66 m/s.
This plot shows us the velocity at inlet is 3.66 m/s and is constant throughout the simulation.
Case 1:
Case 2:
Case 3:
From the above figures we can see that the velocity at outlet is greater than the velocity at inlet. This is due to the decreased density at outlet, pressure at outlet also decreases and thus the velocity increases.
6. Super-cycle parameters: Three cases for supercycle time-length parameters were taken to consideration while comparing it with the baseline configuration. The size of base grid for each was 0.003m
Case 1: Time length for each cycle stage=0.01sec
Case 2: Time length for each cycle stage=0.02sec
Case 3: Time length for each cycle stage=0.03sec
The simulation time marginally increased with the increase of stage time interval for super-cycle parameter.