[don6alfonso]

Hey,
I have a problem and some questions about the libAcoustics library.
I simulate a pipe-flow.

My problem is:
-the the pressure fluctuation calculate but not the fft. I just copy the testcases and modify them for my problem.

My questions:
1. what is the effect of pRef in the observers options?
2. my results CurleAnalogy_1-time.dat are not really different? Is this realistic?
3. What is the meaning of the force which is calculated in the Curle.C? need this more settings?

Here are the files from my case:

application     pisoFoam;
startFrom       latestTime;
startTime       0;
stopAt          endTime;
endTime         0.0011;
deltaT          1e-6;
writeControl    timeStep;
writeInterval   100;
purgeWrite      0;
writeFormat     ascii;
writePrecision  6;
writeCompression off;
timeFormat      general;
timePrecision   6;
runTimeModifiable true;

functions
{

    CurleAnalogy_1
    {
    functionObjectLibs     ("libAcoustics.so");
    
    type                       Curle;
    log                          true;
    probeFrequency        1;
    patchNames              (Wall);
    timeStart                    1e-3;
    timeEnd                     1.1e-3;
    c0                               510;
    dRef                          -1; 
    pName                      p;
    rhoName                  none;
    rhoRef                      0.4097;
    
    observers
    {
        microphone-A
        {
        position    (3.8 0.1 0);
        pRef          2.0e-5;
        fftFreq        1024;
        }
    }
    } 
 
}

I started my simulation with potentialFoam then simpleFoam and after with pisoFoam.

Thank you.
Best, Adrian

[skeptik]

Do you use 2.3 version?

In this case you should compile FFTW lib and ensure that it's working

[don6alfonso]

Originally Posted by skeptik

Do you use 2.3 version?

In this case you should compile FFTW lib and ensure that it's working

I'm using OF 3.0.1.

I think there is it include.

[skeptik]

Originally Posted by don6alfonso

I'm using OF 3.0.1.

I think there is it include.

Yes, it's included but libfftw.so also should present in proper location

[don6alfonso]

Originally Posted by skeptik

Yes, it's included but libfftw.so also should present in proper location

I don't know what you mean with proper location. The information in the controleDict/functions/CurleAnalogy are not enough?

Now its working, I have nothing change only the number of iteration / time steps.
Need the Fourier transform a minimum number of values?

[skeptik]

Originally Posted by don6alfonso

I don't know what you mean with proper location. The information in the controleDict/functions/CurleAnalogy are not enough?

Now its working, I have nothing change only the number of iteration / time steps.
Need the Fourier transform a minimum number of values?

Of course. It's basic thing. May be later we will add some lines to prevent such issues

[skeptik]

Really , we have troubles in SoundObserver::fft

Now, I'm fixing it

[guanjiang.chen]

Originally Posted by don6alfonso

Hey,
I have a problem and some questions about the libAcoustics library.
I simulate a pipe-flow.

My problem is:
-the the pressure fluctuation calculate but not the fft. I just copy the testcases and modify them for my problem.

My questions:
1. what is the effect of pRef in the observers options?
2. my results CurleAnalogy_1-time.dat are not really different? Is this realistic?
3. What is the meaning of the force which is calculated in the Curle.C? need this more settings?

Here are the files from my case:

application     pisoFoam;
startFrom       latestTime;
startTime       0;
stopAt          endTime;
endTime         0.0011;
deltaT          1e-6;
writeControl    timeStep;
writeInterval   100;
purgeWrite      0;
writeFormat     ascii;
writePrecision  6;
writeCompression off;
timeFormat      general;
timePrecision   6;
runTimeModifiable true;

functions
{

    CurleAnalogy_1
    {
    functionObjectLibs     ("libAcoustics.so");
    
    type                       Curle;
    log                          true;
    probeFrequency        1;
    patchNames              (Wall);
    timeStart                    1e-3;
    timeEnd                     1.1e-3;
    c0                               510;
    dRef                          -1; 
    pName                      p;
    rhoName                  none;
    rhoRef                      0.4097;
    
    observers
    {
        microphone-A
        {
        position    (3.8 0.1 0);
        pRef          2.0e-5;
        fftFreq        1024;
        }
    }
    } 
 
}

I started my simulation with potentialFoam then simpleFoam and after with pisoFoam.

Thank you.
Best, Adrian

Hi，

Reference pressure pRef is always set 2E-5. It doesn't need too much change I think.

Do you know what fftFreq means? sample frequency or nfft.

For curle method, I find the number of frequency in the result is the same with fftFreq. But for farassat method, fftFreq seems to have no effects on the result frequency.

Regards,
Guanjiang

[xCFD]

Originally Posted by skeptik

Really , we have troubles in SoundObserver::fft

Now, I'm fixing it

Hello,

i am having the same problem. my case is working fine and i am getting pressure fluc w.r.t. time but fft is not calculated. it does nothing. Did you solve this problem? do you know what we have to do to overcome this problem?

regards,
Ammad

[guanjiang.chen]

Hi,

Which version do you use? I think the latest version has solved this problem. By the way, there is a question panel about libAcoustics

https://github.com/unicfdlab/libAcoustics/issues

Regards,
Guanjiang

[xCFD]

Hello,

i found the solution, actually it was due to fftfreq. i set it to the correct value and now its working perfectly.

Thank you

[guanjiang.chen]

Hi,

What value do you set for fftfreq? Could you share your fwh set file with me.

Regards,
Guanjiang

[xCFD]

Actually i had a look into the source code that what is triggering fft calculations. There i found that the fftfreq should be the value after which you want to calculate fft spectrum. for eg. if you want to calculate every 100 timesteps then set fftfreq to 100.

i hope this helps

[guanjiang.chen]

Thank you for your reply.

I calculate the SPL using the fluctuation pressure in the "xxx-time.dat" file. I don't understand the FFT process of libacoustic very well. Is the fftfreq related to the frequency range of your SPL result? what decides the SPL frequency range, the time steps?

[xCFD]

Originally Posted by guanjiang.chen

Thank you for your reply.

I calculate the SPL using the fluctuation pressure in the "xxx-time.dat" file. I don't understand the FFT process of libacoustic very well. Is the fftfreq related to the frequency range of your SPL result? what decides the SPL frequency range, the time steps?

Yes fft frequency range is calculated by simulation time step. ofcourse if you are simulating with 0.0001 sec of time step then the max frequency your fft can resolve is 10000 Hz.