Arrange the peak magnitude values of multiple strain gages according to their locations.

[vgrchada]

Hi Luis,

Regarding my first problem, I have been doing the first and second methods that you suggested since the beginning of this work. The only problem with those methods is, they are tidious. Every time I have to carefully pick the points from each trial, note it down and then assign it in the Labview. There are numerous trials that I have to study. The third and fourth methods that you suggested are nothing but the questions/problems that I asked. The fourth method that you suggested is exactly what I am looking for.

Regarding my second problem, there are resampling.vi's available in signal operations of full version LabVIEW. I gone through each of them but doen't seem to be useful to solve my problem. I will download the free version of advance signal processing tookit and look in that. I discussed about this problem with my colleagues and they suggested that there could be some thing in wavelets that can solve this problem. What do you say about that? By the way, you said before that you got the strain energy graph by integrating between two adjacent data points. Can you discuss how to do that in labview?

Thank you

Vgrchada

[vgrchada]

Dear Luis and Alex,

I found something interesting. The integration for finding area of graph.vi gives the area of the graph for the assigned threshold limits on x-axis. In my case, the area between two thresholds (i.e., the adjacent points of x-axis on the MS (XY graph)) should be repeated every time it takes the following adjacent points and that area should be represented as a bar in the SE (XY graph). The final SE (XY graph) can be obtained by joining the bars (i.e., area over each zone) with the spline curve.  I don’t understand how he got 5 bars in between every 2.5 spacing on x-axis. I need that because the number of bars should be as many as possible to capture the complete information from the MS (XY graph).  

The other vi gives the different ways to solve the integration.

Thank you

vgrchada

[Tortu]

Greetings Vgrchada.

Regarding the selection of proper data, let me explain a little more in depth the third and fourth methods. The third method is simply assigning a constant value for the start and end rows that will contain any actual range of data. So for example, if you set Start as row 6 and End as row 1200, if the data is actually between 640 and 845 or 500 and 900, it won’t matter, since you will be reading it anyway. After that, just use the Max and Min function to get the actual value that you are interested. The fourth method is just writing somewhere in the spreadsheet file a cell with the actual start and end rows values. How this cell is written, and how will the actual row values be determined is up to you. That is, it will probably require that you write it yourself in the spreadsheet file. But, every time the program runs, make it so that it reads the specific cell that you arbitrarily wrote to (I suggest using the exact same cell for all files) and from that information, set the start and end row for the rest of the program.

Now, about the integration; you mentioned that you need to increase the number of bars between the points to increase accuracy. That is what the resampling is for. Right now, your samples are around 20 inches apart. If you use resampling to modify the “sample rate” of your acquired signal to 1 inch increments, you will be able to have more points in you integration VI’s input and improve your accuracy. My guess is that your process involves interpolation for this very same reason. Since the signal acquisition is discrete (only 32 points in the continuum of the structure) interpolation will allow generating a more fine grain “acquisition”, moving the model closer to the continuous realm, allowing integration of the signal.

Do you have any other questions? Let us know. Have a good day.

[vgrchada]

Hi Luis,

Thanks for your suggestions on both the issues. I will try to take care of the first issue myself but I need more guidence on second issue. I understand clearly what you said about both the issues and the suggestion on the second issue is exactly what I am looking for. Can you walk me through those steps to find the integration of the signal? Moreover, I need to be able to zoom in the signal to capture as much information as I can by using scaling and wavelet functions. The studies need to be more precise in my case because the further studies in civil engineering are depended on this. From now on, I will be waiting for your replies and respond to you as soon as possible because the deadline is close for this issue. If you want me to do this in the support service by calling you, I can do that too or if you want to continue in the discussion forum, I am fine with it. I just want the issues to be solved. Let me know.

Thank you

Venu

[Tortu]

Greetings Vgrchada

The VI I use for the example that I attached belongs to the Time Series Analysis palette, that comes with the Advanced Signal Processing Toolkit I believe. There are other resampling VIs available on other palettes, it is just that I personally like this one. The important thing to notice is the result of using a resampling VI. Notice the increase of samples after changing the “sample rate” of the signal. Connecting the signal to any integration VI will give you the results that you need.

Let me know if you need more help with this issue.

[vgrchada]

Hi Luis,

I got it right but just needs a small correction. The first strain gage is located at 26inches from the support. So the figure has to be movee to a distance of 26inches from the start point and the resampling has to be inbetween 0 and 26inches also.

Then in the next step, I tried with different integration.vi's but the requirements are different for each of them and I don't understand what the small red dot that appears after wiring the cluster to the integration.vi and I think that is not allowing the results to appear. May be I am totally wrong in wiring to the right point. Can you tell me how to get that done.

Thank you

Venu

[Tortu]

Greetings Vgrchada.

The red dot is call a “coercion dot” In this document there is more information on it. (Relevant info is near the bottom of the page.)

https://www.ni.com/docs/en-US/bundle/labview/page/using-wires-to-link-block-diagram-objects.html

Now, it is most likely that the integration VI is getting confused, so in order to recover the actual data from the waveform type, you can use the “Get Waveform Components” that is under Programming>Waveform palette. The “Y” is the array data.

Let us know if the issue is solved

[vgrchada]

Hi Luis,

I did what you said. The coercion dot is gone because of coverstion to similar data types of input. It make sense that the Y in the get waveform component is only selected and sent for the integration but now the integration results are just the strainght line graphs which tells that there is something wrong again. I don't understand which integration.vi can perform integration zone by zone and how. That is the main issue.

Thank you

Venu

[Tortu]

Greetings Vgrchada

This VI should do the trick. In the resampled signal, the “dt” should be 1 inch (at least in my example, I don’t know which resampling you applied).

At this time, the problem has moved too far away from the original issue and I recommend starting a new thread or calling National Instruments for a more agile support.

Have a good day.

[vgrchada]

Hi Luis,

I used the same resampling.vi that you used in your example and I tried the integration.vi before and tried it one again after you suggested me. The result is a numeric integer but not a 1D array to plot the graph. The graph that should look just same as the dotted curve in the photo 4, which you have already seen before.

That other small correction that I need is in the graph obtained after resampling.The first strain gage is located at 26inches from the support. So the figure has to be move to a distance of 26inches from the start point and the resampling has to be in between 0 and 26inches also.

I am sorry for not making the subvi's of my program. I learning it just now and should be done with that soon. Thanks for bearing with me. I don't want to explain all this again from the beginning to other person to resolve the issues. So, once these issues are resolved. we will close this discussion.

Thank you

Venu