Changing Frequency causing Error? Answered

Hi everyone.

I am wondering why the arrival times for my signals change when I change the frequency. Everything else is kept the same and I am only changing the excitation frequency.  I am simulating through Titanium Grain Structure using water boundaries with absorbing boundary conditions.

I am using a Ricker wavelet and as you can see from my image as I increase frequency, the arrival times shorten. Not sure why this would happen. Also, my front wall signals should have the same amplitude as when leaving the excitation point, the wave is only travelling through water until it reaches the front wall.

Another problem is the amplitude of my back wall signals are increasing with frequency when I theory they should decrease. I have tried to account for damping in my Titanium microstructure and the water. Just wondering if anyone would have an Idea why I seem to be getting backwards results.

Thanks :) 

6 comments

  • Hi David,

    I'm not sure that your plots are showing different travel times. From the images you provided it looks like the signals have different widths which is expected since they have different frequencies. What can be see clearly is that the pulses end at different times but it is not clear if the start of the arriving pulses occur at different times. Please check the difference between the zero crossing after the peak of the first and second pulses for each of the frequencies to confirm.

    Can you also some more context and detail regarding your later questions. I think the image below relates to those but I'm not sure how.

    Thanks,

     

    Comment actions Permalink
  • Hey Tapiwa,

    Thanks for the response. 

    From the image below, I have zoomed in on the very first signal which is the activation signal from where I am exciting my wave. You can see the  different widths of each signal (Red 10 MHz, Blue 7 MHz and Black 5 MHz) and  from the zero crossings after each peak we have different arrival times? Or am I interpreting this incorrectly ?



    Regarding the later questions. As I increase the excitation frequency I would expect the amplitude response from each successive back wall to decrease.  I have put a plot of my Excitation frequency Vs Amplitude below. You can see that my front wall amplitude decreases as frequency increase but this could be due to beam divergence??

    The results you can see from my back wall, I would expect this to decrease, not increase. Not sure if there are other parameters within my model I am not considering, but the main parameter I am thinking is crucial is getting the damping correct. I have tried to add this into my model, not 100% sure if I have correctly but I am still getting the same results. 

    Thanks :) 





    Comment actions Permalink
  • Hi David,

    Thanks for sending the zoomed image, I think it shows that the pulses start at different times as measured from the first zero crossing after the peak. The same process can be done for the first echo where you take the first zero crossing after the peak for the echo pulse to determine the arrival time. The difference between arrival time and start time should give a better estimate of flight time.

    With regards to the amplitude response, this response is complex can be variable depending on your setup. It is reasonable to expect the amplitude to decrease, but only if we are working in the far field. Within the nearfield, the amplitude can be much more variable. As you change the frequency, the radiation characteristics of your source change as well so please be mindful of these factors.

    Thanks,

    Tapiwa

    Comment actions Permalink
  • Hi Tapiwa,

    Thanks for your response. The nearfield - far field response is something I genuinely haven't considered.  Thanks for this, something I can read into.

    I was more worried that my modelling was incorrect. I have been building up this model for some time now and can't seem to find anything wrong except trying to add in the damping. Another one of your colleagues has provided me with matlab scripts that plots the damping of the model based on input parameters. So the lines of damping I have included in my model. I can not say with full confidence that I have executed this correctly or getting the correct results. 

    Thanks :)

    Comment actions Permalink
  • Hi David,

    Thanks for clarifying. It is challenging to determine what can be going wrong especially when several parameters can influence the characteristic of interest. To avoid the radiation profile influencing your response, you can generate plane waves to isolate the effect to material properties. It's a good check for attenuation. I am attaching some scripts which use a plane wave model to infer attenuation properties, they might be helpful as a reference. I will send these separately.

    Thanks,

    Tapiwa

    Comment actions Permalink
  • Hi Tapiwa

    Got them and thanks again. This is very appreciated.

    Thanks

    Dave :)

    Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post