[Omish]

Hi everyone.
Pressure drop in my case should be about 25 kPa. instead the pressure contour show an amount of 5 kPa which is not acceptable at all. what could be the problem?
I calculated the h_l parameter and then dP and it was about 10 kPa. so even if we don't consider h_m (for elbow) it's still too low.
The case a U-shape copper pipe with 0.53 m/s velocity inlet, with laminar flow.

[raghulvr]

what function are you using to compute pressure drop values?
use vertex average function in the fluent menu to calculate pressure drop.

[Omish]

Originally Posted by raghulvr

what function are you using to compute pressure drop values?
use vertex average function in the fluent menu to calculate pressure drop.

This is the contour for static pressure

So there wouldn't be any much difference by using some other ways to calculate, and since it's reigons are perpendicular to pipe, pressure drop can be easily computed by subtracting the max & min of the contour.

[raghulvr]

i think the pressure drop value of 25KPa for this problem is very high and surely this will not be the value for this setup.i request you to verify the pressure drop values.
Moreover,in my experience for a exhaust muffler system the pressure drop value does not exceed 4-5KPa.In your case,that too for a 0.53m/s flow velocity,surely this will not be the value.


yes you are correct.u can verify this value by going to ansys cfd-post.if u define the expression
areaAve(Pressure)@REGION:__________ where u want to compute the value of pressure drop

[Omish]

Originally Posted by raghulvr

i think the pressure drop value of 25KPa for this problem is very high and surely this will not be the value for this setup.i request you to verify the pressure drop values.
Moreover,in my experience for a exhaust muffler system the pressure drop value does not exceed 4-5KPa.In your case,that too for a 0.53m/s flow velocity,surely this will not be the value.


yes you are correct.u can verify this value by going to ansys cfd-post.if u define the expression
areaAve(Pressure)@REGION:__________ where u want to compute the value of pressure drop

The 25 kpa pressure drop is gained by an experimental study. I know that results of a 2D modeling would have much difference with experimental results. But not as much as 20 kPa difference is acceptable!
I used the h_L = f(L/D)(V^2/2*g) and then h_L = (delta P)/Gamma
To calculate the major loss. (Gamma is the specific weight of the fluid)
The result was about 10 kPa. So even if we don't consider the loss because of two bends , 5 kPa is still very low even compared with theoretical answer.
I attached my calculation in case you want to check. And here are the data I used:

Q=9.072718 (l/min)
D=19.05 (mm)
density= 865.93 (kg/m^3)
dynamic viscosity = 0.108 (pa.s)

Sorry I made a mistake with the word. Yes I meant "region". And by that I just meant the color boundaries which show areas with different pressures.

[raghulvr]

i have used ANSYS CFX to run my simulations.but i also tried with fluent and got absurd results like these.so i request you to use cfx for this study.
i am sure you will get your result validated with cfx.just give it a try

[Omish]

Originally Posted by raghulvr

i have used ANSYS CFX to run my simulations.but i also tried with fluent and got absurd results like these.so i request you to use cfx for this study.
i am sure you will get your result validated with cfx.just give it a try

Thank you so much for your help Raghulvr. I would try CFX and let you know about the result.
I was wondering if you could find the problem with Fluent and tell me about it. I would be really thankful if you do this. Thanks again.

[raghulvr]

i have also tried to figure out what the problem is but i really don't know .my prof recommended me to use cfx unless the problem has some multiphase or heat transfer in it.so in general for fluid flow problems use cfx
also define the pressure drop function like this.
1.define two boundaries for those you want two monitor and calculate pressure drop.for ex.inlet and outlet.
2.use shear stress transport model.
3.define a expression
areaAve(Pressure)@REGION:INLET-areaAve(Pressure)@REGION:OUTLET

[Omish]

Originally Posted by raghulvr

i have also tried to figure out what the problem is but i really don't know .my prof recommended me to use cfx unless the problem has some multiphase or heat transfer.so in general for fluid flow problems use cfx

I found a problem :-(
CFX can't solve 2D models. What could I do?

[raghulvr]

if you want to have a 2D simulation, you have to draw a geometry in 2D and with a small thickness ( this geometry is in 3D but solved in 2D). define both side of thickness as symmetry. See tutorial 7 of CFX tutorial 11.0 (free surface flow over bump)

[Omish]

Originally Posted by raghulvr

if you want to have a 2D simulation, you have to draw a geometry in 2D and with a small thickness ( this geometry is in 3D but solved in 2D). define both side of thickness as symmetry. See tutorial 7 of CFX tutorial 11.0 (free surface flow over bump)

I'm new to Ansys. Could you explain with more details? Where should I define symmetry for planes? Is it ok if I use Design modeler > tools > symmetry? Or it should be done in CFX itself?
I found the tutorial you mentioned but it's a bit confusing. I don't know where to start.

[raghulvr]

symmetry can be done only in design modeler.create a 1mm thick section for your geometry.thats enough to change the object from 2d to 3d

[Omish]

Originally Posted by raghulvr

symmetry can be done only in design modeler.create a 1mm thick section for your geometry.thats enough to change the object from 2d to 3d

I tried to do that.In design modeler after I activate symmetry, faces of the geometry can not be selected !

[Omish]

Originally Posted by raghulvr

symmetry can be done only in design modeler.create a 1mm thick section for your geometry.thats enough to change the object from 2d to 3d

I tried to create two new planes on both sides. They could be selected now but after clicking "Generate" it gives an error:
"Symmetry feature failed"
I guess it's because the planes do not intesect with my geometry. Shouldn't I select only one plane and that is the middle plane instead?

[Omish]

Originally Posted by raghulvr

i have used ANSYS CFX to run my simulations.but i also tried with fluent and got absurd results like these.so i request you to use cfx for this study.
i am sure you will get your result validated with cfx.just give it a try

Hi
I also used CFX for my simulation and the results were wrong like fluent. Any suggestions?
Here's my theoric calculation for pressure drop. I've checked it so many times and don't see any problem in it.

[liliana]

Originally Posted by Omish

Hi
I also used CFX for my simulation and the results were wrong like fluent. Any suggestions?
Here's my theoric calculation for pressure drop. I've checked it so many times and don't see any problem in it.

I have the exactly same problem! For my cases, my calculations results in 0.16 Pa of pressure drop, but Fluent gave me 0.08 Pa. My calculations were equal to yours. Have you found a solution to this problem?

[hamed.majeed]

Originally Posted by liliana

I have the exactly same problem! For my cases, my calculations results in 0.16 Pa of pressure drop, but Fluent gave me 0.08 Pa. My calculations were equal to yours. Have you found a solution to this problem?

What are the boundary conditions you use? How do you initialize the solution??

[liliana]

Originally Posted by hamed.majeed

What are the boundary conditions you use? How do you initialize the solution??

Hi!

At inlet, a parabolic velocity profile with vmax = 0.016 m/s
At oulet, pressure-outlet Gauge Pressure = 0

The geometry has 0.2m diameter and 1m length (2D).

The flow is laminar.

Rho = 827 kg/m³
Viscosity = 0.026 Pa.s

By hand, I calculate dP = rho * (L/D) * (64/Re) * ((v_ave)^2)/2 and got 0.163 Pa. With Fluent, I got 0.082 Pa.

So I got around a half of it.

[liliana]

Originally Posted by hamed.majeed

What are the boundary conditions you use? How do you initialize the solution??

I initialized using "hybrid initialization" with the options on default. I did not change anything on Fluent.

[hamed.majeed]

Originally Posted by liliana

I initialized using "hybrid initialization" with the options on default. I did not change anything on Fluent.

You can try this:
Hybrid Initialization>More Settings>General Settings>Initialization Options> Use specified Initial Pressure on Inlets: Select it.

Send me your case file if that doesn't work!