[ANSYS Meshing] Simple heat exchanger meshing

Scaty · October 26, 2021, 04:56

Hi everybody,

I tried to set up a fairly simple case in ANSYS Workbench with the solver Fluent.
Later on I will solve a heat exchanger, but as this is quite perfomance-demanding I carry out the meshing-study with a simplified model to find out which meshing I should use in the later full model to get a reliable result.

My model is a 2D rectangular (0.04 m wide and 1 m long) with a velocity inlet on the left side and a no-pressure outlet on the right. The long walls are my boundaries and both are set to no-slip and a fixed temperature of 20 °C. My inlet temperature is 80 °C and the velocity is high enough to ensure a turbulent flow (Reynolds number above 10000, I use the k-epsilon turbulence model).

Now about my actual question. Every time I change the meshing before running the solver, my results change significantly. At first when the mesh is rather coarse that is plausible as such a coarse mesh is presumably not able to solve the boundary layers on both sides. But even if I refine my mesh or add inflation layers at the edges, or even if I generally increase the number of cells, my results keep changing (sometimes with fine meshes they even diverge?). Usually I expect only slight alterations in the solution when I reached the point on which the mesh is good enough, but that never happens.

Logically I am mostly interested in the temperature profile and the average outlet temperature, but the other values like pressure and velocity keep changing as well.

Is there anything I am doing wrong, any ideas how I can resolve that issue?
I am happy about help, it is quite frustrating to spend hours and hours without any progress

.

mluckyw · October 27, 2021, 05:21

What Mesh shape do you use? I think for your case with a rectangular geometry would have more benefit by using a structured hexahedral mesh or in case of 2d QUAD surface mesh. How about the mesh quality? is it acceptable?
Have you tried by using another turbulence model such as SST and keeping the Y+ value to an appropriate value for each turbulence model?

October 26, 2021, 04:56	Simple heat exchanger meshing	#1
Scaty New Member Join Date: Oct 2021 Posts: 3 Rep Power: 5	Hi everybody, I tried to set up a fairly simple case in ANSYS Workbench with the solver Fluent. Later on I will solve a heat exchanger, but as this is quite perfomance-demanding I carry out the meshing-study with a simplified model to find out which meshing I should use in the later full model to get a reliable result. My model is a 2D rectangular (0.04 m wide and 1 m long) with a velocity inlet on the left side and a no-pressure outlet on the right. The long walls are my boundaries and both are set to no-slip and a fixed temperature of 20 °C. My inlet temperature is 80 °C and the velocity is high enough to ensure a turbulent flow (Reynolds number above 10000, I use the k-epsilon turbulence model). Now about my actual question. Every time I change the meshing before running the solver, my results change significantly. At first when the mesh is rather coarse that is plausible as such a coarse mesh is presumably not able to solve the boundary layers on both sides. But even if I refine my mesh or add inflation layers at the edges, or even if I generally increase the number of cells, my results keep changing (sometimes with fine meshes they even diverge?). Usually I expect only slight alterations in the solution when I reached the point on which the mesh is good enough, but that never happens. Logically I am mostly interested in the temperature profile and the average outlet temperature, but the other values like pressure and velocity keep changing as well. Is there anything I am doing wrong, any ideas how I can resolve that issue? I am happy about help, it is quite frustrating to spend hours and hours without any progress .

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Gmsh] gmshToFoam generates patches with 0 faces and 0 points	Simurgh	OpenFOAM Meshing & Mesh Conversion	4	August 25, 2023 08:58
Apply heat sink/mimic heat exchanger in closed loop with buoyantBoussinesqPimpleFoam	alex1001	OpenFOAM Running, Solving & CFD	0	October 28, 2019 14:38
[snappyHexMesh] meshing the geometry of an heat exchanger	RobSie	OpenFOAM Meshing & Mesh Conversion	0	November 30, 2016 11:22
[Gmsh] Vertex numbering is dense	KateEisenhower	OpenFOAM Meshing & Mesh Conversion	7	August 3, 2015 11:49
Convective Heat Transfer - Heat Exchanger	Mark	CFX	6	November 15, 2004 16:55

October 27, 2021, 05:21		#2
mluckyw Member Chrowale Join Date: Sep 2021 Location: Bandung, Indonesia Posts: 63 Rep Power: 5	What Mesh shape do you use? I think for your case with a rectangular geometry would have more benefit by using a structured hexahedral mesh or in case of 2d QUAD surface mesh. How about the mesh quality? is it acceptable? Have you tried by using another turbulence model such as SST and keeping the Y+ value to an appropriate value for each turbulence model?