Mastering ANSYS: Propeller noise in Fluent (aeroacoustics)

[@chris57926]

Great tutorial! Thank you!

[@caewitharminhashemi]

@@chris57926 you're welcome

[@RobineSabat]

Hello, thank you for the tutorial. I made the same instructions as in the video but in the configure part I see that the interface between the two cylinders is defined as walls. What can be the problem please?

[@caewitharminhashemi]

@@RobineSabat hello, you're welcome. In mesh, describe section, you have to set fluid fluid interfaces to no walls. Watch 5:20

[@RobineSabat]

@@caewitharminhashemi Yes, I chose yes for "Change all fluid-fluid boundary types from wall to internal", and I still have the same problem :(

[@caewitharminhashemi]

@RobineSabat change walls to interface from setup

[@RobineSabat]

@@caewitharminhashemi Thank you so much. I appreciate it!

[@caewitharminhashemi]

@RobineSabat happy to help

[@JjownO]

I would like to ask how to set the height of the first boundary layer of the blade? I saw that the first layer height and number of layers of the boundary layer in your two videos are different.

[@JjownO]

Three warnings appeared when I generated the volume mesh. Warning: region fluid Warning: Skipping synchronization as either size field is not available or size field is already synchronized. Warning: Min. aspect ratio less than 1 at 484 nodes

[@caewitharminhashemi]

From my experience, a height between 2 and 4 micrometers is proper. After the solution, you have to check y plus. It should not exceed 2.

[@caewitharminhashemi]

Decrease the minimum size of the surface mesh.

[@JjownO]

@@caewitharminhashemi Thank you very much!

[@nidhijapillay8044]

Hi, is it possible to do prediction of noise of subsonic get from nozzle using the similar setup as in the video?. Thank you for the video. Why did you not choose to use FW-H surfaces for this model

[@caewitharminhashemi]

Hello. If the propeller tip Mach is low. No need to define a surface (permeable FWH) because the quadrupole sound source is weak. You have to define surface in jet noise calculation because quadropole source is dominant

[@nidhijapillay8044]

@@caewitharminhashemi Thanks a lot for your explanation.

[@arezoonajafian797]

Excellent I didn't understand what you put values on inlet and outlet boundary conditions? Can i do it for my axysymetry propller by one blade??

[@caewitharminhashemi]

Inlet and outlet are pressure outlet. You can add a inlet velocity if the propeller moves in air.

[@caewitharminhashemi]

No. You cannot use symmetry or periodic boundary condition in mesh motion. It uses in stationary solution.

[@arezoonajafian797]

@caewitharminhashemi thank you🙂‍↕️ So I can't simulate the noise of one blade.🥲 All of it has to be.

[@franekknecht6991]

Hi, I seem to have some error in my solution. When i plot pathlines of velocity, there is a sudden change in particle trajctories. before mrf region my trajectory coincides with axis of proppeler's rotation, subsequently it experiences a swirl inside MRF region, and then suddenly while exiting the MRF trajectory suddenly straightens. Do you happen to have an idea where the root of the error might be

[@caewitharminhashemi]

Hello. Please email me some photos of path lines. [e-mail]

[@Aarones-wv6hf]

Thank you for the tutorial. This is really helpful. When I'm processing the final data I get a message saying 'Data set x-axis is not in increasing order' when I click on fast fourier transform. I did increase the number of time steps but I doubt that's what's happening

[@caewitharminhashemi]

Hello. Unfortunately, I have not encountered this error yet.

[@BhaveshTavte]

how did you set the co-ordinates for the recivers to locate over the propeller can you help me sir or anyone who knows that thing

[@caewitharminhashemi]

There is a circle with a diameter of 2 meters, which is far from the center of the propeller.

[@aryanmantrawadi8571]

heyy, what should be the model parameters of far-field density and sound speed in water in the acoustic model

[@caewitharminhashemi]

Hello, 1500 m/s, density is 1000 and reference pressure sound is 1us

[@muteki1408]

Superb tutorial thank you very much! I have some problem where the SPL results in CFD-post are showing negative values, which is unexpected. My simulation setup is exactly the same as yours and I used the DJI propeller blade as well. I've been searching for solution for quite a while and the only answer I've got is that I have to change the reference pressure. So I tried 3 different reference pressure: 0 Pa, 20 micro Pa and 1 bar but the SPL results are still negative. Could you please give me some guidance on this matter? Thank you very much!

[@caewitharminhashemi]

Hello. Check yplus it must be below 2. What is your fluid model?

[@md.adamyamin2643]

I am doing a combustion simulation to analyse thermoacoustics. When I try to read the acoustic signal files I get "Read_Next_Acoustic_Timestep: Error opening input files!". Why do we get this error? Also, is it necessary to get a converged solution before applying Acoustic model?

[@caewitharminhashemi]

Hello, what are fluid and acoustics model?

[@JjownO]

Why can't I set the buffer layer when generating the volume grid ？No buffer layer setting in my fluent!

[@caewitharminhashemi]

Did you use body of influence?

[@JjownO]

Yes!

[@caewitharminhashemi]

@@JjownO You cannot use both. Remove boi to access buffer layers

[@JjownO]

@@caewitharminhashemi thank you!

[@beingahmed276]

Great Tutorial...Great Teacher...You should look to create courses on platform alike Coursera/Udemy etc.... Mind sharing the cad model??

[@caewitharminhashemi]

Thank you. I'm only a simple engineer. I downloaded this propeller from grabcad website. I don't remember exact name of model. As I know it's DJI propeller.

[@adityaslab]

Enjoy

[@JjownO]

In the transient calculation process, my convergence is very fast and very stable, but when the calculation is more than 80%, the residual suddenly rises tens of thousands of times, and the monitoring of the rotor thrust value also changes to tens of thousands N. Do you know the reason🤔

[@caewitharminhashemi]

It happens sometimes. Because mesh motion is used. In the steps interval that diverges, increase the residual value. After passing that interval, return it to the first values.

[@JjownO]

@@caewitharminhashemi Thank you for your reply. When I met this situation for the first time, your way can solve the problem effectively. But when I ran into the problem the second time after 60 more steps, I tried to raise the residuals to 0.1,10,100, 10,000, but still couldn't diverge after calculating up to 12 steps. And the residual value is getting larger and larger.🥺

[@caewitharminhashemi]

@@JjownO check mesh and time step. problem is due to one of these

[@oy-5742]

请问，你知道如何计算螺旋桨展向的升力分布和升力系数分布吗？

[@caewitharminhashemi]

Hello. I think this video could help you. kzbin.info/www/bejne/rZPOqHakdt-UaJosi=jm-vPVMRfx5WjRwp

[@박진경-u7h]

I want to know the same method at student license. I'm following the video but "existing due to license issue"

[@caewitharminhashemi]

The student license has a limited cell number. Increase mesh size

[@박진경-u7h]

​@caewitharminhashemi I already know about the cell limits. how do I increase the size? I'd like to know the exact size figures as shown in the video above.

[@caewitharminhashemi]

@@박진경-u7h If you are a student, use an academic account

[@박진경-u7h]

​@caewitharminhashemi so there is no ell limit for academic account

[@caewitharminhashemi]

@@박진경-u7h It is less restrictive.

[@aleyraihan]

do we have to always use the steady solution first or we can just go with transient without steady ?

[@caewitharminhashemi]

The steady solution increases the speed of convergence. If you use the transient solution directly, it may diverge.

[@aleyraihan]

Thank you for your reply, i still have more question, what if i add local sizing for the volume of the propeller instead of the rotational zone ?, does it effect the accuracy of the result or is the result is still viable because my machine is struggling with the rotational zone local sizing while if i instead use the local sizing only on the propeller volume, i can run the analysis easily

[@caewitharminhashemi]

@@aleyraihan There's no exact rules for mesh size. You cannot predict results by one solution. You have to solve a solution by three different size of mesh, coarse, medium and fine. Yplus must be below 2 for LES and DES and between 30 and 60 for rans models.

[@aleyraihan]

Okay i’ll try it again, but problem is, i can’t use the exact mesh setting because it takes too long for my machine. I am only able to go as low as 0.2mm for the mesh because less than that will take too long (>1 day) for the simulation. I’m trying to find a viable solution that doesn’t need high computational power

[@aleyraihan]

@@caewitharminhashemi So, first, i have to thank you so much for this tutorial, and i comment because i have done all the step until CFD post, after fast fourier transfrom, i got a spl vs frequency charts that show ups and down line of the chart, Is this normal ?, and also because ups and downs of the chart, i also got negative SPL, is this right ? i placed the receiver 1m in front of the propeller. Thank you for guiding me until this far

[@shuratomt]

my solution is diverging and reverse flowing, and I have no clue why

[@caewitharminhashemi]

It depends on several parameters. First you have to check the quality of mesh.

[@JjownO]

the following prompt appears When setting the transient solution Warning: explicit time stepping not available Using 1st order dual time🤔

[@caewitharminhashemi]

Yes. If you want to use LES and DES models you have to change it to bounded second order. I u want to use rans models, you have to change it to second order.

[@JjownO]

​@@caewitharminhashemi Thank you! Your video and answers are very clear, thank you very much!

[@caewitharminhashemi]

@@JjownO you're welcome. Good luck