Taking simultaneous readings from strain gauges, lvdt's and loacells

[sfjd]

Hi,
I have PXI-1052, PXI-6251, three SCXI 1520, three 1314 terminal blocks, one SCXI 1540 and one 1315 block. In experiment I have the following devices:

    six strain gauges, six lvdt's (three in x direction and three in y direction)and four load cells.

I would like to take readings from strain gauges and lvdt's at selected load increments.  Also I would like to see the plot of load versus deformation  (i.e. lvdt average values in x and y direction each) and control my application of load accordingly.

I have tried with MAX, but what I found with my little knowledge that it is giving an output in a text file with x value as time and y value as the strain, displacement, etc. How do I integrate the above operation so that I can see the behaviour in realtime?

Regards,
Rajeevan

[Brian F.]

Hello Rajeevan.

Thank you for contacting National Instruments.

I am a bit confused as to how you are logging your data to a text file in MAX.  Are you using VI Logger or VI Logger Lite to accomplish this or are you accomplishing this in some other way? 

As far as seeing your data real-time, you will need to pursue another path to do this in addition to writing to a text file.  This is because when the text file is being written to, it cannot be accessed by any other program or user of the computer.  To work around this, you could feed the data into a graph (in MAX or LabVIEW) or a numeric indicator (in LabVIEW) so you could view this data in real-time as well.

If you wish to vary what is graphed on the x axis and the y axis, the best way to accomplish this is the 'XY graph' control in LabVIEW

Let me know if this answers your question.  If you have further questions, we would be glad to help!

Brian F

Applications Engineer

National Instruments

[sfjd]

Dear Brian,
Thank you very much for the prompt reply. You are correct. I was using the VI logger option inside MAX to dump the file (it is availabel inside MAX). I have also tried with L:ABVIEW 7 to create a DAQmx task to view the data in real time. But my problem is diffrerent:

FIRST PROBLEM:

I created a DAQmx task and created one loadcell and one strain gauge task. Then I have used a waveform graph indicator to plot the data. What happened is both loadcell and strain gauge readings are shown in the same graph. As you are aware both scales should be different, since one is loadcell and the other is strain and there is no sense in viewing them in the same scale and graph.

QUESTION:
    If it is possible to create multiple task under same DAQmx assistant VI, how to view them in different graph indicators? That is one graph for strain gauge and one for load cell. I also tried to create one more DAQmx assistant task inside the same VI, but only one (either strain or loadcell) is running.

SECOND PROBLEM

In my experiment, as I mentioned earlier, I have 6 strain gauges, 4 loadcells(two load cell each on either side of the specimen applying tension), and 6 lvdt's (three horizontal and three othogonal to it in a horizontal plane). What I want to do is
    (i) for each increment of load,say 10 kN, I would like to take strain gauge and lvdt readings.
    (ii) to display in graphical form  the load(average of either pair of load cell readings say at 0, 10, 20, 0 kN intervals) versus  deformation      (average of three lvdt's in x directon) at the corresponding load level. This is to get an idea about the variation of load versus deformation in while the experiment is progressing and control the load application.

What is the best way to do it?

Regards,
Rajeevan

Message Edited by sfjd on 12-29-2006 11:04 PM

[sfjd]

Hi,
I have also tried to plot the load vesus strain(for the time being i don't have lvdt connected; so I am using strain data) using XY graph. The vi is enclosed. I did not see anything in the graph. The DAQ assistant is configured to have two channels to read loadcell and strain.

I have also included an error report which I get while running the VI.

Regards,
Rajeevan

Message Edited by sfjd on 12-30-2006 06:12 AM

[devchander]

Rajeevan,

In continuous mode, you have selected Samples to read as 10 and scan Rate as 1K.

Here, you are losing out on acquired samples and hence the error

So, either increase the samples to read ( For ex: like 100 or 500) or decrease the scan rate to avoid the error that you are currently getting

Message Edited by devchander on 01-01-2007 04:39 AM

[sfjd]

Hi Dev,

Thanks for the reply. Now I face a different problem. When  I configured through "custom voltage with excitation" one load cell, I got the following problem:

 Data:

 HBM C6A  1 MN (equal to 100 000kg(f) tonne) load cell. Sesitivity 2mV/V. Full bridge

I entered + 6m and -6m under the settings tab max and min range; Excitation 3 V(Since excitation is 3 V, I entered 3 * 2mV/V = +-6m in the input max and min range). I did not opt for any scale.

Under device tab, I checked Auto zero once and low bypass frequency as 10 kHz(I don't know what to give under the lowpass filter frequency option, so i selected 10 kHz).

When I clicked test and applied 100 tonne through a UTM, I got the output as 6.87 mV as output. Almost 0.87 * 100/6~ 14 tonne excess. I also used the calibration option under the third tab in the same window, in which  I increased my load through 10 tonne interval. In none of the case it showed the correct value almsot varied by 2 tonne to 5 tonnes.

Can you please help?

Regards,

Rajeevan

[Brian F.]

Hello sfjd. 

I think part of the problem resides with your minimum and maximum value selection.  They are currently configured for +-6 mV, but the value you are expecting from the 100 tonne measurement is also 6 mV.  Thus, any amount over this value will be railing the voltage.  I would recommend increasing these limits and see if this helps you get more accurate readings.  Secondly, I would recommend calibrating the load cells for the full range that you wish to measure over.  This will correct any error in the readings and should help a more accurate measurement to be taken. 

Good luck on your application!

Brian F
Applications Engineer
National Instruments

[sfjd]

Dear Brian,

Thanks. I tried to calibrate it using the calibrate command button under the third tab after device tab of the configuration window.( I created a DAQ assistant inside labVIEW 7.1 and it popped up the configuration window.). Is it the proper method of calibrating it or else is there any other method?

Second question is if I keep my range more than 6mV how is it corresponding to my load cell? For example, my load cell give a sesitivity of 2 mV and when I gave 3 V excitation, the maximum expected is 6 mV not more than that. Any value more than that means the load cell is in error or 2 mV/V sensitivity is incorrect?

Suppose if I created a custom voltage with excitation task for a load cell with a custom made scale:

As I metnioned earlier my load cell is 100 tonnes.

In the maximum I gave, say, 150 tonnes which corresponds to (6/100)* 150 = 9 mV. Will the load cell output that much higher voltage?

Also If I am giving a map range to read the calues in tonnes than in volts, what shoul I enter in the prescaled max and min columns?

Like these:

    Prescaled:  Max = + 9 mV and Min -9 mV
   Scaled: Max = +150 tonnes Min: -150 tonnes

Is it necessary that I give the negative range also like above. Can I give 0 to +150 tonnes also?

What about filter frequency? And Autozero mode?

What is this Autozero mode will do? When ever I select Autozero mode before selscting any measurement, the first reading at zero load is not indicated as zero instead it is showing some value around 0.8 tonnes.

And also one more question. I created an LVDT (poistion) task. I tried to calibrate it. with two values (low and high), In the calibration table it showd me zero and 2 mm correclty. But when I save the calibration and clicked the test button, for 2 mm displacement it showed only 1.852 mm. How is this difference happening?

Regards,
Rajeevan

[Brian F.]

Hello Rajeevan.

You are correct in the way that you are doing calibration via MAX and this should be enough to give you the performance that you need. 

The range set by configuring the Max and Min Signal Input Range parameter effectively allows greater resolution on your card over a smaller range.  You may choose to have a minimum value that isn’t the maximum value multiplied by negative 1 (i.e. you don’t have to have 5 and -5, it can be 5 and -4). 

It is important to make sure that the 2 mV/V configuration is accurate since this is your scaling factor.  Even though you may be reading completely valid values pre-scale, if the scaling is off, you will then see incorrect values post-scaling.

Auto zero mode is a procedure for eliminating offsets generated by an amplifier stage.  More information on how to use this feature can be found in the following two knowledge base articles.  They also have great information on the general process of setting up a transducer measurement in MAX. 

Measuring a Full-Bridge Absolute Transducer with the SCXI-1520 or PXI-4220

http://digital.ni.com/public.nsf/allkb/fd9f22608fdc939a86256b5a0075705d

Removing Large Initial Offset for Load Cell or Strain Measurements with SCXI-1520 or PXI-4220

http://digital.ni.com/public.nsf/allkb/DA2ABCFA3D4C73FD862571080008B8B1

Filter frequency allows you to place a filter on the data to eliminate high frequency data that is not expected to come from a transducer measurement. 

As far as your question regarding your LVDT, how are you controlling the fact that the displacement is only 2 mm?  If you extend the measurement to a larger distance, does your accuracy increase?  Or do you still see the 7.5% error in measurement?  Please also refer to this knowledge base article for some extra information on using the 1540 with an LVDT. 

Using the SCXI-1540 with a Device that Requires External Excitation

http://digital.ni.com/public.nsf/websearch/5FB0106294CE483886256A02006091A5?OpenDocument

Finally, what type of chassis are you using?  Are you using a combo chassis?  Are there any other SCXI modules in the chassis along with the 1520 and 1540’s?

Have a great day and let us know how your application progresses.

Brian F

Applications Engineer

National Instruments 

[sfjd]

Dear Brian,

The 2 MV/V sensitivity is specified by the load cell manufacturer HBM in their data sheet(Load Cell type C6A). When I check the Autozero mode to Once and low pass filter frequency to 1 kHz, I get for no load a load of about 850 kg(f). When the load is increased to  10 tonnes, it showed 11.2 tonnes, for 20 tonnes 21.2 tonnes, etc. But for 50 tonnes it showed 50.3 tonnes and for 100 tonnes it showed 99.8 tonnes. It means that initially there is a difference of about 1.2 tonnes but at some intermediate value the error  is negligible. How is this happening. My measurement requires to apply only about 10 tonnes of load ( I don't have any load cell with low range that is why I am using 100 tonnes). Since the DAQ output shows such erroneous results how can I depend on this measurement ?

The chasis is PXI -1052, DAQ is multifunction M series 6251, SCXI-1520 and SXCI 1314, SCXI 1540 and SCXI 1315.

The problem is that when I repeat the calibration I am not able to reproduce the same value which I got previously. I am quite frustuated with this measurement. What could be the problem? Is it a problem of the UTM(Universal Testing Machine) which applies the load, or the problem of the load cell, or te problem of the indicator used to read the load from UTM (it is digital display indicator from HBM) or the problem of NI? I could not sort it out yet?

Also I have an experimental programme for which I wrote a VI in LabVIEW 8, which needs to be run continously for two hours. I have 19 channels to read and this I embedded in DAQ assistant.  Will there be any memory problem. Since I see my excel file 4 MB for running almost one minute I don't have any idea about whether the program will crash or not? Anu syggestions?

A signal splitter is used to read individual signals and display. I would like to see the load versus lvdt readings in a graph like display. One problem is that it shows only connecting straighlines rather than a smooth curve. I tried single acquistion mode as well as continuous mode. But of no use. I also used time delay of one second inside the block diagram. Sampling rate 1000 and frequency 1k. Any help?

REgards,

Rajeevan