Mixing axisymmetry and convective boundary condition

MV78 · July 20, 2018, 06:40

Hi,
I am working on a shell-and-tube heat exchanger. The model is 2D axisymmetric. I'd like to model the heat transfer from the inner tube with a convective boundary condition.

Without a convective boundary condition, the axisymmetry axis corresponds to the bottom side of my rectangular domain. However, when I represent my fluid with a convective boundary condition the domain is reduced. The bottom side corresponds now to the convective boundary condition and I can no longer define an axisymmetry axis as it is out of the boundary limits. Creating a separate line disconnected from the mesh isn't possible.

I could use a planar symmetry to get around this problem but it will introduce errors. Does anyone know how I could proceed ?

Thank you in advance,
MV78

LuckyTran · July 20, 2018, 10:43

In 2D the axis for the axisymmetric condition is always the x=0 line. You only need the axis BC when you have a boundary on the axis (e.g. in a pipe). For an annulus I think you don't need to do anything.

MV78 · July 20, 2018, 11:19

Of course. Let's say that the axis is the x=0 line. I want to define a convective boundary condition for the line x=0.01 and there is no domain between the convective BC and the axis. However, my problem is that the axis has to be a boundary condition of a domain to exist.

I may be unable to see the solution as I am working with Hypermesh for the mesh.

LuckyTran · July 20, 2018, 21:07

I'm not sure why you actually need to specify an axis because the imported mesh is a 2D rectangle or something. It sounds like you have a hypermesh problem and not a Fluent problem.

A possible workaround though is to generate your 2D rectangle using an axis at x=0. Import this mesh into Fluent. Then translate it upwards along y to x=0.01 m.

MV78 · July 21, 2018, 07:57

The rectangular geometry represents the 2D axisymmetric model of the shell-and-tube heat exchanger. I don't have a problem to generate the axis along the x=0 line, I can create a perfect axisymmetric model. The issue arises when I want to simplify my model with a convective BC at x=0.01 instead of the hot transfer fluid. I can't specify an axis at x=0 as it isn't a domain boundary anymore, the one from the former hot transfer fluid domain.

LuckyTran · July 21, 2018, 17:39

You don't need to specify an axis at x=0, the axis is always at x=0. You only use the axis BC when you actually have a boundary at x=0, because otherwise it would be missing a BC. If you do not have a boundary at x=0, then you don't need to do anything more. You only need to specify that it is a 2D axisymmetric type simulation in the setup.

July 20, 2018, 06:40	Mixing axisymmetry and convective boundary condition	#1
MV78 New Member Join Date: May 2018 Posts: 8 Rep Power: 9	Hi, I am working on a shell-and-tube heat exchanger. The model is 2D axisymmetric. I'd like to model the heat transfer from the inner tube with a convective boundary condition. Without a convective boundary condition, the axisymmetry axis corresponds to the bottom side of my rectangular domain. However, when I represent my fluid with a convective boundary condition the domain is reduced. The bottom side corresponds now to the convective boundary condition and I can no longer define an axisymmetry axis as it is out of the boundary limits. Creating a separate line disconnected from the mesh isn't possible. I could use a planar symmetry to get around this problem but it will introduce errors. Does anyone know how I could proceed ? Thank you in advance, MV78

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

July 20, 2018, 10:43		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,837 Rep Power: 68	In 2D the axis for the axisymmetric condition is always the x=0 line. You only need the axis BC when you have a boundary on the axis (e.g. in a pipe). For an annulus I think you don't need to do anything.

July 20, 2018, 11:19		#3
MV78 New Member Join Date: May 2018 Posts: 8 Rep Power: 9	Of course. Let's say that the axis is the x=0 line. I want to define a convective boundary condition for the line x=0.01 and there is no domain between the convective BC and the axis. However, my problem is that the axis has to be a boundary condition of a domain to exist. I may be unable to see the solution as I am working with Hypermesh for the mesh.

July 20, 2018, 21:07		#4
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,837 Rep Power: 68	I'm not sure why you actually need to specify an axis because the imported mesh is a 2D rectangle or something. It sounds like you have a hypermesh problem and not a Fluent problem. A possible workaround though is to generate your 2D rectangle using an axis at x=0. Import this mesh into Fluent. Then translate it upwards along y to x=0.01 m.

July 21, 2018, 07:57		#5
MV78 New Member Join Date: May 2018 Posts: 8 Rep Power: 9	The rectangular geometry represents the 2D axisymmetric model of the shell-and-tube heat exchanger. I don't have a problem to generate the axis along the x=0 line, I can create a perfect axisymmetric model. The issue arises when I want to simplify my model with a convective BC at x=0.01 instead of the hot transfer fluid. I can't specify an axis at x=0 as it isn't a domain boundary anymore, the one from the former hot transfer fluid domain.

July 21, 2018, 17:39		#6
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,837 Rep Power: 68	You don't need to specify an axis at x=0, the axis is always at x=0. You only use the axis BC when you actually have a boundary at x=0, because otherwise it would be missing a BC. If you do not have a boundary at x=0, then you don't need to do anything more. You only need to specify that it is a 2D axisymmetric type simulation in the setup.