[cfd_de]

Hello,

i was looking on the forum on how to visualize the reflection of pressure wave from a wall boundary. Some people suggested that it can be visualized in the CFD Post with the help of pressure contours. Can someone please explain with an example how it can be visualized ? Like what's the indication ? Can someone show some picture ?

At the wall boudary the flow will come to rest and there will be a pressure rise. Is this an indication of reflection of pressure wave from a wall ?

Looking forward for some help.

Regards

[ghorrocks]

If you put a plane through the domain and colour the plane by pressure you should see the wave reflecting. If the wave is oblique to the surface you should see the reflection, if it is parallel to the surface you might need a transient animation to see it.

[Opaque]

If you look at

https://hal.archives-ouvertes.fr/hal-00802480/document

you will find two use cases to study the reflective properties of discretization scheme at a wall boundary.

Keep in mind ANSYS CFX does not use the scheme described in the paper; however, the verification test proposed is universal and should help answer your question.

[cfd_de]

Originally Posted by ghorrocks

If you put a plane through the domain and colour the plane by pressure you should see the wave reflecting. If the wave is oblique to the surface you should see the reflection, if it is parallel to the surface you might need a transient animation to see it.

Thanks for your reply.

I have attached an animation which i made in CFD Post. Sorry its a fast one because i couldn't find out how to control the speed of the animation in CFD Post.

Can you see some wave reflecting from the cylinder boundary ? The contour lines are moving back and forth very often. So what's the clear indication of wave reflecting from a wall boundary ?

[cfd_de]

Sorry I was not able to attach the .zip file which contains the animation file because its size exceeded the maximum allowed size. Is there any other option to attach large files ?

[cfd_de]

Originally Posted by Opaque

If you look at

https://hal.archives-ouvertes.fr/hal-00802480/document

you will find two use cases to study the reflective properties of discretization scheme at a wall boundary.

Keep in mind ANSYS CFX does not use the scheme described in the paper; however, the verification test proposed is universal and should help answer your question.

Thanks for your reply.

I will go through this paper but is this contains some information to visualize the pressure wave reflection from a wall boundary ?

[ghorrocks]

Attachments: https://www.cfd-online.com/Wiki/Ansy...n_the_forum.3F

Do not attach an animation as it will probably be too big. Just post some still images from it which show what you are talking about.

[cfd_de]

Originally Posted by ghorrocks

Attachments: https://www.cfd-online.com/Wiki/Ansy...n_the_forum.3F

Do not attach an animation as it will probably be too big. Just post some still images from it which show what you are talking about.

Please see the images in the attachments.

Some introduction of problem:
Inlet : Pulsating Mass Flow B.C- its only positive half of sine wave which then repeats itself

Outlet : Moving Cylinder- Modelled as Rigid Body with specified Force in the Y-direction

Just beyond the Inlet, the flow gets divided into two paths which then meets again just before the cylinder

Compressible Flow Simulation

0.140825 [s] marks the begin of a new Mass flow cycle and the time step is chosen such that one cycle is divided into 50 steps.

I am posting just 15 images from one cycle, at the alternate steps as can be seen from the time step selector in the attached images.

[cfd_de]

Images 6-10

[cfd_de]

Images 11-15

[Opaque]

Originally Posted by cfd_de

Thanks for your reply.

I will go through this paper but is this contains some information to visualize the pressure wave reflection from a wall boundary?

As Glenn indicated, if you want to see pressure wave being reflected, you may need a transient animation, in particular a waver normal to the boundary.

Say you want to visualize the reflection at the outlet, you can setup a simple 1D duct with slip walls, set a sinusoidal total pressure intent BC, and static pressure outlet BC. Run the simulation long enough for the wave to reach the outlet plus a bit more to see the reflection return.

Now, the post-processing:
1 - Set static pressure monitor points at a different location along the duct
2 - You should see the pressure pulse build-up going from inlet to outlet
3 - Once the pulse is reflected it will show up as a distortion of the original signal established earlier. Usually, it will change the amplitude (addition or subtraction), and it could change the phase a bit.

Usually, these reflections are "relatively small" compared to pressure level, so it is best to look at Pressure - Pressure.Trnavg.. Since you are imposing the wave, you already know the transient average; therefore, the value is fixed (do not use transient statistics or it will take forever to see the true issue)

[cfd_de]

Originally Posted by Opaque

As Glenn indicated, if you want to see pressure wave being reflected, you may need a transient animation, in particular a waver normal to the boundary.

Say you want to visualize the reflection at the outlet, you can setup a simple 1D duct with slip walls, set a sinusoidal total pressure intent BC, and static pressure outlet BC. Run the simulation long enough for the wave to reach the outlet plus a bit more to see the reflection return.

Now, the post-processing:
1 - Set static pressure monitor points at a different location along the duct
2 - You should see the pressure pulse build-up going from inlet to outlet
3 - Once the pulse is reflected it will show up as a distortion of the original signal established earlier. Usually, it will change the amplitude (addition or subtraction), and it could change the phase a bit.

Usually, these reflections are "relatively small" compared to pressure level, so it is best to look at Pressure - Pressure.Trnavg.. Since you are imposing the wave, you already know the transient average; therefore, the value is fixed (do not use transient statistics or it will take forever to see the true issue)

Thanks for the detailed reply.

I want to see the wave reflection from the cylinder boundary, which is modelled as a rigid body so i can't use the static pressure boundary condition in this case.

The monitor points which i would setup to monitor the pressure at different locations along the domain, will be only discrete points. How they will give information regarding the propogation of pressure wave ? Can you please explain a bit more ?

Also Glenn had mentioned that the pressure wave propogation and reflection can be monitored in CFD Post with the help of Pressure plots. I am trying to understand how this would work.

[Opaque]

I think you will benefit by creating a few expressions in CFD-Post for a traveling wave, here is an applet with the basics

https://www.geogebra.org/m/YKuG3zNZ

the applet shows a single traveling wave, you can create a secondary traveling wave in the opposite direction, and added to the first one. You will see then what happens to the original traveling wave and the kind of distortions that will occur

Create a chart for your expressions evaluated at a point, then animate the chart in time.

Now, the same can be done for a surface plot, or contour. Recall the previous advice about removing the level, so you focus on how a perturbation travels.

[cfd_de]

Originally Posted by Opaque

I think you will benefit by creating a few expressions in CFD-Post for a traveling wave, here is an applet with the basics

https://www.geogebra.org/m/YKuG3zNZ

the applet shows a single traveling wave, you can create a secondary traveling wave in the opposite direction, and added to the first one. You will see then what happens to the original traveling wave and the kind of distortions that will occur

Create a chart for your expressions evaluated at a point, then animate the chart in time.

Now, the same can be done for a surface plot, or contour. Recall the previous advice about removing the level, so you focus on how a perturbation travels.

If i use Pressure contour plot on a section plane at different time steps for the evaluation then what exactly would indicate a wave traveling and the wave reflection from a wall boundary ?



I animated the pressure contours over a particular time, the contour lines moves back and forth several times so i am wondering what exactly would indicate a pressure wave ? The movement of pressure contour lines or something else ?



You can get the idea of the movement of pressure contour lines from the pictures attached in the previous replies.

[TurBoris]

Hi,

I have a couple of fundamental questions regarding the wave reflections.

-What is the math behind the reflection from a pressure boundary condition or why does a pressure wave reflect from a pressure BC ?

-Does the reflection process occur in a same way as reflecting from a wall BC ?

-Is the effect of the reflection to inner domain same for both transient and steady simulations ?

-For turbomachinery applications where pressure outlet (or opening) BC is placed close to blade trailing edge (in some cases it is a must), should I turn on the non-reflective BC in CFX with default settings ?

-When the model is steady and includes mixing plane close to shock wave of the blade, is there any treatment in CFX to prevent the reflections ?

Thank you in advance.

[ghorrocks]

I do not know the mathematics of a reflections at pressure boundaries. I just know they do reflect - try it and find out.

It is not the same as a wall reflection. The mathematics there is clear, the velocity is fixed at zero which causes the pressure to rise. And that high pressure then propagates away from the wall as a reflection.

Using a non-reflective boundary near a trailing edge will probably help, but does not guarantee it will work. I am pretty sure it just reduces the reflection, not eliminate it; and if there is a recirculation passing through the boundary it will not work well with that.