CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   wall thinning (http://www.cfd-online.com/Forums/cfx/107909-wall-thinning.html)

natalie October 9, 2012 07:12

wall thinning
 
Hello everyone!

I need to simulate wall thinning (in tube, for example) depending on flow properties with CFX.
Have you got any ideas how to do this?
Should i use moving walls or is there any other way to do this?
if anyone has an experience in solving such type of problems, any advice wuold be a great help.

Thanks in advance

evcelica October 9, 2012 07:33

What do you mean by wall thinning?
Erosion?

natalie October 9, 2012 07:35

Yes, erosion..

evcelica October 9, 2012 07:40

One of the CFX tutorials is an erosion tutorial with sand particles.
I believe it is: Tutorial 9: Flow Through a Butterfly Valve.

natalie October 9, 2012 07:50

Thank you, i've seen all of them already.
I've got no particles in my simulation, so these tutorials and built-in models of erosion don't suite me.
Erosion in my simulation should move the walls depending on flow properties. I mean that the zone of flow expands with time as the walls themselves become thinner.
Hope, i've clearified my problem

Thanks a lot

energy382 October 9, 2012 08:49

Quote:

Originally Posted by natalie (Post 385703)
Thank you, i've seen all of them already.
I've got no particles in my simulation, so these tutorials and built-in models of erosion don't suite me.
Erosion in my simulation should move the walls depending on flow properties. I mean that the zone of flow expands with time as the walls themselves become thinner.
Hope, i've clearified my problem

Thanks a lot

sounds like a fsi problem

natalie October 9, 2012 09:06

Thanks
i've been thinking about this, too..
but maybe just solving flow fields and moving some of the wall nodes according to the predefined law would be enough..?
in fact the behaviour of solid material itself doesn't interst me.. just the interface between solid and fluid

energy382 October 9, 2012 09:21

there's an option to model thin walls. You've to chose a fluid-fluid domain interface with thin material and no slip wall.

Normally, it is used to model heat transfer through thin solids or to represent coatings (fluid solid with thin material option)


So I don't think, it is suitable for your issue.

Ahhh, just forgotten, that there's a thing called mesh deformation. You can describe mesh motion (e.g. nodal displacements) with expressions or functions.

Otherwise, I've no idea how to solve your problems (without a fsi simulation).

natalie October 9, 2012 09:34

What do you mean with "thin material" and "thin solid" for heat transfer??
is it the same thing, that was used in tutorial with can combustor for thin blades?
as i can guess, the main idea there is to model correct heat transfer without using too many mesh nodes&
can this boundary change its position?

thanks

natalie October 9, 2012 09:36

to energy382: "Maybe Glenn has the experience to answer your questions"

I hope so too )))

energy382 October 9, 2012 09:40

Quote:

Originally Posted by natalie (Post 385728)
What do you mean with "thin material" and "thin solid" for heat transfer??
is it the same thing, that was used in tutorial with can combustor for thin blades?
as i can guess, the main idea there is to model correct heat transfer without using too many mesh nodes&
can this boundary change its position?

thanks

in simulations, where boundaries are moved, you can use this mesh deformation option.

natalie October 9, 2012 09:43

is this "thin wall" anywhere described in CFX documentation?
i've never used it before..

thanks for your interest ))

energy382 October 9, 2012 09:45

Quote:

Originally Posted by natalie (Post 385734)
is this "thin wall" anywhere described in CFX documentation?
i've never used it before..

thanks for your interest ))

give me your email (PM), I'll send you some papers

natalie October 9, 2012 09:48

Thank you very much!

my email is: korolkova.natash@yandex.ru

evcelica October 9, 2012 09:49

I'm guessing this wall thinning will take place over a looooooong time period? If so you wouldn't want to try to simulate this transiently. Perhaps you could run a set of steady state runs with the walls "thinned" different amounts, and then calculate the transient thinning from your set of data.
Workbenches input/output parameters may be helpful for this.

natalie October 9, 2012 09:57

yes, for sure, running it with real time - isn't the proper way for this problem..)
you're right with idea of the set of steady-state simulations with changing wall positions and in fact, that's the way probem had been solved until now. But now the man idea is to do it automatically, without interrupting the first simulation.
and the new interface positions are not known apriori because they depend on flow properties


All times are GMT -4. The time now is 13:21.