Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Calculate flux through surfaces in 2D axial symmetry
Posted 14 ott 2010, 03:51 GMT-4 1 Reply
Please login with a confirmed email address before reporting spam
I am trying to calculate the flow/flux of a liquid through a surface in a laminar flow model, but have troubles getting correct values. The model is a simple 2D axial symmetry model. I have a inlet and an outlet (controlling the flow) .
To test how it works I try to calculate the flow using a line integral on the outlet using a selection (choosing only the outlet). With a global evaluation I calculate the following:
w*OutletRad^2*pi
But this will not give me the correct values. If I do
w*OutletRad*pi [m]
I come close, but I don't understand why (as I understand the integral it should normalize for the length of the line it integrates). It seems like the error becomes bigger if i try to do the same evaluation on other surfaces in the model where the flux should be the same (non-compressible flow and pressure constraint)
I have also tried to use the reacf(w) and putting this in a line-integral evaluation and set it to do a surface integral (can do this because of the axial symmetry). This method gives me a flux of 0
Any suggestions or comments to my methods
Best
To test how it works I try to calculate the flow using a line integral on the outlet using a selection (choosing only the outlet). With a global evaluation I calculate the following:
w*OutletRad^2*pi
But this will not give me the correct values. If I do
w*OutletRad*pi [m]
I come close, but I don't understand why (as I understand the integral it should normalize for the length of the line it integrates). It seems like the error becomes bigger if i try to do the same evaluation on other surfaces in the model where the flux should be the same (non-compressible flow and pressure constraint)
I have also tried to use the reacf(w) and putting this in a line-integral evaluation and set it to do a surface integral (can do this because of the axial symmetry). This method gives me a flux of 0
Any suggestions or comments to my methods
Best
1 Reply Last Post 17 ott 2010, 05:23 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
in 2D axial symmetry you see a cut view of a volume that is obtained by integrating along the loop length "2*pi*r".
COMSOL always works in 3D, but by default show you 2D with a default 1[m] depth along Z (out of the paper).
For axi symmetry its the "loop length" that gives the link from 2D to 3D. So when you select an edge in 2D axi you are slecting a surface of int(2*pi*r*your_function,dr) in [m^2].
In V4 COMSOl has changed slightly and now includes the 2*pi see the doc
you can also if you use the "Derived Values - Line integration - Integration settings" check the "compute surface integrals" to automatically add the 2*pi*r.
Note this option is not inlcuded in the integration operators you might define in the Definition node section
The word is always check your integrations on simple cases when you can check the values easily, then go to the complex cases where simple handcalculations is harder to do.
--
Good luck
Ivar
in 2D axial symmetry you see a cut view of a volume that is obtained by integrating along the loop length "2*pi*r".
COMSOL always works in 3D, but by default show you 2D with a default 1[m] depth along Z (out of the paper).
For axi symmetry its the "loop length" that gives the link from 2D to 3D. So when you select an edge in 2D axi you are slecting a surface of int(2*pi*r*your_function,dr) in [m^2].
In V4 COMSOl has changed slightly and now includes the 2*pi see the doc
you can also if you use the "Derived Values - Line integration - Integration settings" check the "compute surface integrals" to automatically add the 2*pi*r.
Note this option is not inlcuded in the integration operators you might define in the Definition node section
The word is always check your integrations on simple cases when you can check the values easily, then go to the complex cases where simple handcalculations is harder to do.
--
Good luck
Ivar