fem/application of the load due to fluiddynamic resistance of shape

[Simsto]

Good morning to all,
I find myself asking for help from all of you about a probably trivial problem but cause confusion.
I have to carry out by fem a linear static analysis of a circular cylinder immersed in a pipeline run by steam overheated.
the cylinder is positioned as a bounce, with its axis positioned orthogonally compared to the steam flow lines.
My problem arises in attributing the correct loads of force and/or pressure on the outer surface of this cylinder.
I know that the viscosity of the steam fluid would tend to drag the cylinder if it was not bound by the structure.
the forces at stake therefore are due to a difference of pressure between the semicircumference invested by the upstream fluid and the downstream fluid.
speculating that the steam pressure is 8bar, I calculated that the p-resistant is approximately 1bar (using as a cross section of the cylinder).
which of the various assumptions is correct?
1) impose a pressure on the semi-face upstream (8+1)bar and on the semi-face downstream 8bar
2) impose a pressure on the semi-face upstream of (1)bar and on the semi-face downstream 0bar
3) impose on the half-face up a pressure of (8)bar and on the semi-face downstream (8-1)bar
that is, the pressure deriving from the resistant force is to be set in addition to the hydrostatic one of the upstream steam or to be subtracted in the downstream one?

thank you very much in advance, if you need more details the delucidations

 

[Onda]

Good morning to all,
I find myself asking for help from all of you about a probably trivial problem but cause confusion.
I have to carry out by fem a linear static analysis of a circular cylinder immersed in a pipeline run by steam overheated.
the cylinder is positioned as a bounce, with its axis positioned orthogonally compared to the steam flow lines.
My problem arises in attributing the correct loads of force and/or pressure on the outer surface of this cylinder.
I know that the viscosity of the steam fluid would tend to drag the cylinder if it was not bound by the structure.
the forces at stake therefore are due to a difference of pressure between the semicircumference invested by the upstream fluid and the downstream fluid.
speculating that the steam pressure is 8bar, I calculated that the p-resistant is approximately 1bar (using as a cross section of the cylinder).
which of the various assumptions is correct?
1) impose a pressure on the semi-face upstream (8+1)bar and on the semi-face downstream 8bar
2) impose a pressure on the semi-face upstream of (1)bar and on the semi-face downstream 0bar
3) impose on the half-face up a pressure of (8)bar and on the semi-face downstream (8-1)bar
that is, the pressure deriving from the resistant force is to be set in addition to the hydrostatic one of the upstream steam or to be subtracted in the downstream one?

thank you very much in advance, if you need more details the delucidations

but the fluid at what speed is it? is it not that perhaps the cylinder undergoes a function force of speed and resistance coefficient? ? .
the pressure has little to do, apart from the fact that the density varies.
therefore, the cylinder, in its total, undergoes a force equal to: 1/2 ro a v^2 cd.
where with cd you have the strength coefficient of a cylinder, function of the number of reynolds. the density of steam, ro, depends on the pressure.
then, as you want to apply this force on the cylinder, depends on many factors, if the motion is laminating, turbulent, etc. but for the analysis of the cylinder itself is usually of little importance.
all this if the inside of the cylinder is at the same pressure as the outside. If instead it is at a different pressure, the pressure difference must be applied on the faces.

So, with regard to the model, I would apply, if there is a difference of pressure between inside and outside, a constant pressure on all faces (equal to the delta between inside and outside). more pressure (which in the first approximation can be constant) on the invested face and a pressure contrary to the underwind face, such as to generate a force equal to that indicated by the quoted formula.
with regard to some documentation, due to the fact that the flow around a cylinder is complicated and dictated by different phenomena according to the number of reynolds:https://www.unirc.it/documentazione/materiale_didattico/599_2010_264_8138.pdf https://moodle2.units.it/pluginfile.php/185787/mod_resource/content/1/flusso_cilindro.pdf then, if you want to go deeper, in the above links you will find the pressure trend according to the circumferential position. then, you can apply a variable pressure according to the data provided. but it is usually useless, unless the cylinder wall is so thin that it is influenced by local pressure values.

 

[Simsto]

but the fluid at what speed is it? is it not that perhaps the cylinder undergoes a function force of speed and resistance coefficient? ? .
the pressure has little to do, apart from the fact that the density varies.
therefore, the cylinder, in its total, undergoes a force equal to: 1/2 ro a v^2 cd.
where with cd you have the strength coefficient of a cylinder, function of the number of reynolds. the density of steam, ro, depends on the pressure.
then, as you want to apply this force on the cylinder, depends on many factors, if the motion is laminating, turbulent, etc. but for the analysis of the cylinder itself is usually of little importance.
all this if the inside of the cylinder is at the same pressure as the outside. If instead it is at a different pressure, the pressure difference must be applied on the faces.

So, with regard to the model, I would apply, if there is a difference of pressure between inside and outside, a constant pressure on all faces (equal to the delta between inside and outside). more pressure (which in the first approximation can be constant) on the invested face and a pressure contrary to the underwind face, such as to generate a force equal to that indicated by the quoted formula.
with regard to some documentation, due to the fact that the flow around a cylinder is complicated and dictated by different phenomena according to the number of reynolds:https://www.unirc.it/documentazione/materiale_didattico/599_2010_264_8138.pdfhttps://moodle2.units.it/pluginfile.php/185787/mod_resource/content/1/flusso_cilindro.pdfthen, if you want to go deeper, in the above links you will find the pressure trend according to the circumferential position. then, you can apply a variable pressure according to the data provided. but it is usually useless, unless the cylinder wall is so thin that it is influenced by local pressure values.

hello wave,
Thank you so much for the answer. I attach a simplified scheme of what I thought to do and that fortunately you have largely confirmed me.
in substance I will apply:
- internal pressure on the cylinder
-external pressure to the cylinder
- force generated by steam flow on the really invested faces
- force generated by the reaction of the water coming out of the cylinder by spray (which I had forgotten in last post).



I saw you advised me to use two forces, one for the invested face and the other for that underwind.
If I use a single one on that invested, do I make a considerable mistake?
to give two numerical values:
external pressure about 100bar
Fluid resistant pressure 4 bar
internal pressure 103 bar

Thank you very much
simone

 

[stevie]

hello wave,
Thank you so much for the answer. I attach a simplified scheme of what I thought to do and that fortunately you have largely confirmed me.
in substance I will apply:
- internal pressure on the cylinder
-external pressure to the cylinder
- force generated by steam flow on the really invested faces
- force generated by the reaction of the water coming out of the cylinder by spray (which I had forgotten in last post).



I saw you advised me to use two forces, one for the invested face and the other for that underwind.
If I use a single one on that invested, do I make a considerable mistake?
to give two numerical values:
external pressure about 100bar
Fluid resistant pressure 4 bar
internal pressure 103 bar

Thank you very much
simone

you don't understand how you calculated the overpressure of 1 bar on the upside.. .
Have you considered approximate formulas for the cylinder? careful that the drag coefficient is strongly dependent on the number of reynolds.
according to me the ideal approach should be that fsi (structure fluid interaction).
by means of a numerical simulation set the boundary conditions and the conditions of the undisturbed steam flow, calculate to the algorithm the distribution of pressure on the cylinder, hence the resulting of the forces and then do the fem analysis on the resultant, and then see how to proceed.
is it an academic or real problem? brutally said there would be an infinity of parameters to be evaluated (speed of fludium, viscosity, viscosity models, turbulence, possible detachment of vortices, etc.).

 

[Simsto]

you don't understand how you calculated the overpressure of 1 bar on the upside.. .
Have you considered approximate formulas for the cylinder? careful that the drag coefficient is strongly dependent on the number of reynolds.
according to me the ideal approach should be that fsi (structure fluid interaction).
by means of a numerical simulation set the boundary conditions and the conditions of the undisturbed steam flow, calculate to the algorithm the distribution of pressure on the cylinder, hence the resulting of the forces and then do the fem analysis on the resultant, and then see how to proceed.
is it an academic or real problem? brutally said there would be an infinity of parameters to be evaluated (speed of fludium, viscosity, viscosity models, turbulence, possible detachment of vortices, etc.).

Hello, Steve.
thanks for the post.
I did not specify how I calculated the overpressure because I could go off-topic since my problem is how to correctly apply some values and not how I found the same, as for the purposes of the application...the value for itself is quite influential (I think).

the case is a real case, which is why the client non it required no cfd analysis and it is impossible unfortunately to have the pressure distribution along the walls the cylinder through a coupled calculation.

to answer your question, given the not perfectly cylindrical conformation of the model and according to the number of reynolds, I used a coefficient of resistance cd=1 in a cautionary way (although in reality it could be even a little less), so I used an overpression exactly of 0.5*rho*(v^2) by steam parameters.

I repeat again the final question of my previous post so that the thread of speech is not lost too much.

in substance I will apply:
- internal pressure on the cylinder
- external pressure to the cylinder
- force generated by steam flow on the really invested faces
- force generated by the reaction of the water coming out of the cylinder by spray (which I had forgotten in last post).
correct?

Thank you very much

 

[Simsto]

Forgive the new post, but if giving some indicative values of boundary condition can be useful there are no problems.
I didn't go too specific to not divage, but I am happy to delve into the issue even "numerically" if it was for everyone interesting.

 

[Onda]

but the cylinder models with shell or solids?
if with shell, you can safely apply the pressure difference between internal and external. If with solids, apart from the compression that sees the material, which is normally influential, you can continue to charge it with the pressure delta.
Then I don't understand how to apply f2. In theory, that should come out of the fact that your cylinder is laundry. So he's got a balance of forces. as, if it were closed, the net result of the forces would be nothing. but not closed, it is with a push equal to the internal pressure (or better to the internal pressure - external) multiplied by the hole area.
so, I would put the force due to the cd of the cylinder in the fluid. we call the aerodynamic force . and the force due to delta p, which as mentioned above, is applied on an unclosed surface.
with regard to the application of the aerodynamic force, this can be applied on the survivor face, at first approximation,. but if you want to see it correctly, part of this force is an overpressure on the survivor face and part is a depression on the underwind face, so, according to the accuracy of the analysis and what you want to achieve, you can go in detail in the application of force. as a flow distributes pressure on the cylindrical surface, depends on the number of reynolds, and a polar diagram is provided in the documents I attached to you previously.
personally, if the objective is to verify the resistance of the cylinder, which base sees a stinging moment, I do not think it is necessary to go to model the distribution of pressure. if instead you are afraid of effects of local instability, then it would be more appropriate to go to properly model the pressure due to aerodynamic resistance.

 

[Simsto]

very exhaustive answer thank you very much.
the scheme presented is a graphic simplification, in reality the "hole" in the cylinder has a noozle where the water comes out as a spray. f2 I calculated it knowing density, speed and exit section of the spray (known data) and its intensity is the pure dynamic component of reaction to the mass of water coming out of the cylinder.
for the rest all very clear, as you hypothesized the end of the simulation is the mechanical resistance of my model with an accuracy therefore not too fine.

if someone wanted to do practice and make a cfd simulation to make the test of the nine is well accepted .:d:d jokes apart thanks for the info.