Poincare Map

[Arthuruuuuuu]

Hello guys,

thanks for your attention. I am acquiring a displacement signal (dime dependent) and making it derivatives (velocity) with an express sub-VI. Very simple.

Howerver, i have to record and plot this two values (displacement and velocity) every each specific time. How can I detect the frequency, divide the period into a integer number "N"  and record the values every time N is equal to 1 or 2 or 10 ....

Actually, my frequency is changing with time also. 

Can any one help me ???

Thanks a lot,

Arthur. 

[Ben]

---

@Arthuruuuuuu wrote:

Hello guys,

thanks for your attention. I am acquiring a displacement signal (dime dependent) and making it derivatives (velocity) with an express sub-VI. Very simple.

Howerver, i have to record and plot this two values (displacement and velocity) every each specific time. How can I detect the frequency, divide the period into a integer number "N"  and record the values every time N is equal to 1 or 2 or 10 ....

Actually, my frequency is changing with time also. 

Can any one help me ???

Thanks a lot,

Arthur. 

---

Feel free to add some more words to help people understand were you need help.

How did you jump from displacement and velocity to frequency?

But the "every time N is equal to 1 or 2 or 10 " is particularly weird.

Ben

[Arthuruuuuuu]

Hello,

I am acquiring a signal of a displacement of a tip freed end of a beam. Whit this signal, I use derivative subVI to get velocity. 
My signal os displacement and obviusly velocity has a frequency. My main VI needs to identify this frequency and devide this period into a integer and finit number N.

Have a look at the figure attached. The blue signal was devided in a such number of intervals. i need to record and plot the values of displacement and velocity every time my signal pass through onde of those specific division. e.g. N = 2.

[Bob_Schor]

Random thoughts:

• While (sadly) largely true, not all of your readers are "guys".
• I like the concept of a "dime-dependent displacement signal".  I'm guessing it has a radius of about 9 mm?
• You absolutely do not want an Express VI anywhere near this (not-completely-understandable) project!
• When you say that you are acquiring a displacement signal, does that mean that DAQmx is somehow involved?  [Ah, that's where the Express VI, the Dreaded DAQ Assistant, and probably its Evil Twin, the Dynamic Data Wire, are involved ...  Begone!]

So much is unclear (especially your "invisible" code, and please attach actual VIs, not just pictures of parts of them).

Have you had any Signal Theory classes?  Do you have a good understanding about how (and why!) taking derivatives adds noise to a signal?  Do you understand the concept of "phase" of a signal?  I suspect you have a couple of Engineering Classes to pass ...

Bob Schor

P.S. -- the Mathematician in me was drawn to your post's title, which seems to have absolutely no relationship to the post!  Make your title relevant.  And better than "Please help" or "LabVIEW Problem".

[Darin.K]

Let's see if we can tighten up some vocabulary so we are on the same page and can focus on your issue.

It sounds to me like you are doing a Poincare reduction for a cantilever oscillator given an initial displacement.  You are measuring the displacement x(t), presumably at some regular interval dt.  You then use an Express VI, for better or worse, to make a finite-difference approximation to the velocity v(t).  Now, both x(t) and v(t) are periodic with some period T and frequency (1/T).  However, due to the changing Young's modulus, that frequency is going to be changing as the oscillations decay.  This changing frequency is likely your problem.

What you want in order to make a Poincare map in this case is to make your x(t) measurement at a point of fixed phase which can be tricky when the frequency is changing.  Do you need to do this in real-time or can you oversample your data and then analyze afterwards?  Some real x(t) data would be useful to get an idea of how quickly the frequency is changing.

Am I getting it so far?

[Ben]

A phase-locked loop may be helpful?

Ben

[Arthuruuuuuu]

Thanks for your answere, Darin.

You are totally right ! Its exactly this that i want to do but I dont know how because time is not controled by me. I dont know how to have access to time information in labview. Or if you may sugest another way to solve this problem. What would you tel me ?

Thanks again,

Arthur.

[Ben]

---

@Arthuruuuuuu wrote:

Thanks for your answere, Darin.

You are totally right ! Its exactly this that i want to do but I dont know how because time is not controled by me. I dont know how to have access to time information in labview. Or if you may sugest another way to solve this problem. What would you tel me ?

Thanks again,

Arthur.

---

While waiting for Darin try...

Block diagram "ctrl space"

Enter Phase

You should see an option for a Phase Locked Loop.

Drop an instance of same on the diagram and read the help.

When I was working in the condensed matter physics for Dr. Jeremy Levy see here he was developing a variation on a AFM called ANSOM where the vibrations of the probe where being used to image the structure of solids.

Then wait to see what Darin can offer.

Ben

[Darin.K]

---

You are totally right ! Its exactly this that i want to do but I dont know how because time is not controled by me. I dont know how to have access to time information in labview.

---

A few points I need to have cleared up before moving on.  You said in the original post that you are acquiring the data, that to me means that you are in fact "in control" of the time.  You would choose a sampling rate and that determines the time information.  

Is this a homework project?  Totally fine if it is, we just need to know up front.  I am getting a sense that you are being given a time series of points and being asked to create a plot.

Ben totally read my mind about the PLL, a very nice method when the frequency shifts are relatively slow compared to the primary frequency.  I need to see data before deciding.

Another method is to find fixed points on the curve (maxima, minima, zero crossings) and use those as references.  But, you may need to interpolate values, if your sampling rate is fixed but the frequency is changing, point N=2 or whatever in one cycle will correspond to a different phase in a different cycle. 

To really help, you need to describe quite clearly and explicitly how you are getting the data, and what the data consists of.  Posting a sample would be most helpful.

[Arthuruuuuuu]

Dear Darin, thanks very much again,

 "You would choose a sampling rate and that determines the time information."

Sure, I have set my sampling rate, but how can I make a kind of division into a samplig period ?

this a homework project?  Totally fine if it is, we just need to know up front. 

This is the beggining of the experimental part of my PhD research into energy harvesting using piezoelectrical materials and shape memory alloy. I have an electromechanical shaker wich provides mechanical vibration to an aluminum cantlever beam. At the free and of the beam, i got a displacement transducer by laser. I am using DAQ and NI9220 to aquire the signal from the trasducer.

Another method is to find fixed points on the curve (maxima, minima, zero crossings) and use those as references. 

This may not help because the main point is: divide the signal into periods, and divide the period into small intervals. Than, measure the displacement and velocity at this moment every period. I am really not sure but dont think this is the path.

Sorry for not post the VI. I am living quiet far from university and its on computer. I will only have access next monday. Soon, will post here.

Thanks,

Arthur.