Dynamic Signal Acquisition

cancel
Showing results for 
Search instead for 
Did you mean: 

Negative results with strain gauge

Solved!
Go to solution

When I run my VI  the results are always negative. I am using the NI9237 along with the NI9945. I have wired my setup as a quarter bridge. There are three wires coming from the strain gauge. I have gone over the wiring and I believe it is correct per NI specifications. Is there anything in the MAX I should be looking at. Not sure why the values are negative.

thanks

Harry Stone

0 Kudos
Message 1 of 6
(20,353 Views)
Solution
Accepted by topic author harry-stone@idexx.com

Hi Harry,

 

There are a few things I would like to clarify: 

- Tensile strain is positive and compressive strain is negative, which is described from a high level in the tutorial linked below. 

Measuring Strain with Strain Gages 
http://zone.ni.com/devzone/cda/tut/p/id/3642 

As compressive strain is negative, you would see negative results within MAX if your strain gauge is experiencing any compression. Please keep in mind an offset may be associated with each transducer, which is why some sensors use a Certificate of Calibration. This is produced by the manufacturer and is included with the sensor as it is sensor specific. The sensor is passed through a testing process to determine its actual response versus the ideal. In this case a table scale can be created to incorporate these values. 

How Do I Make a Custom Scale in Measurement & Automation Explorer (MAX)? 
http://digital.ni.com/public.nsf/allkb/3F6558112FD2C776862575B5004F7F87?OpenDocument 

Not all sensor manufacturers provide a Certificate of Calibration. Either you can create your own table by placing known amounts of strain, force, etc on the sensor and map it to corresponding voltage or you can create a linear scale within MAX adjusting the y-intercept (b) to the necessary value to remove any offset. 

You are using the NI 9237 which supports offset nulling. An offset null is performed with the sensor attached with no load placed on the sensor. Effectively a voltage measurement is taken and this value is subtracted off every subsequent measurement hence removing the initial offset. This takes the place you creating a linear scale and doing this manually. 

The two links below illustrate how to use a custom scale created in MAX within LabVIEW, as well as, coding the custom scale within LabVIEW to remove the MAX dependency. 

DAQmx Acquire with Custom Scale 
http://decibel.ni.com/content/docs/DOC-3706 

Create Custom Linear Scale for Each AI Channel in LabVIEW Using DAQmx 
http://decibel.ni.com/content/docs/DOC-11136

 

 I recommend using a Strain task. The input parameters obtain the information about your strain gauge necessary to perform the conversions to strain. There is a Strain measurement example within the LabVIEW Example Finder (*Open LabVIEW* Help>>Find Examples) designed specifically for the NI9237 that incorporate the offset nulling and shunt calibration features. If you do not have the external wires connected for shunt calibration as referenced in this document  , then you will receive an error. Below is an explanation of Shunt Calibration taken from the NI DAQmx Help (Start>>All Programs>>National Instruments>>NI-DAQ>>NI-DAQmx Help) to help better explain this feature. 

Shunt Calibration (Gain Adjustment) 
You can verify the output of a bridge-based measurement system by comparing the measured bridge output with a calculated value if the physical load on the sensor is known. NI-DAQmx can then use the difference (if any) between the calculated and the measured values as a gain adjustment factor for each measurement. You can simulate applying a load to the bridge by connecting a large known resistor in parallel with the bridge. This resistor, called a shunt resistor, offsets the zero voltage of the bridge. Because the value of the shunt resistor is known, you can calculate the physical load corresponding to the voltage drop of the resistor. 

Use the DAQmx Perform Shunt Calibration VI/function or the DAQ Assistant to perform a shunt calibration, which sets the gain adjustment for a virtual channel. NI-DAQmx then uses this gain adjustment when scaling readings from the bridge. Some National Instruments products include internal shunt resistors. 

This may seem like information overload but I wanted to provide you a detailed explanation for your understanding in addition to the immediate answers. As a synopsis, I would recommend using the 9237 Strain example and utilize offset nulling. Negative values are expected for compression and positive for tension. The How-To Guide below provides an excellent overview of strain gauges, which also includes a video. 

Strain Measurements with Strain Gages: How-To Guide 
http://zone.ni.com/devzone/cda/tut/p/id/7130

 

Hopefully this helps!

Regards,


h_baker
National Instruments
Applications Engineer
Message 2 of 6
(20,340 Views)

Thanks for the help. I have a vi that now does measure the strain and I used the convert strain gauge reading vi to convert from milli volts to microstrain units. The issue I am now having is the conversion function outputs the reading as a positive or (tensile) strain. The strain I am applying is compressive and I believe should show me a negative strain. Before I  used the convert strain gauge reading vi my output was negative but in negative milli volt units. The polarity was correct I just wanted to have it output in microstrain units. It appears the convert strain gauge reading vi puts the output into positive results. Originally when I had the millivolt signal going into a write to measurement file function the output did remain negative. So is there something I can do to the convert strain gauge reading vi that will show the results to show strain as tensile (positive) or compressive (negative.) The reason I ask this is we may be doing testing where I do not know what the strain is and by viewing the output it will tell me whether it is tensile  or compressive.

Thanks

HS

Message 3 of 6
(20,303 Views)

Hello, 

 

For the strain measurement you are taking, when the the specimen is in compression the strain gauge will lower in resistance which will give you a positive voltage across the bridge (Quarter Bridge configuration), this positive voltage (Vsg, the input to the Convert Strain Gauge Reading VI ) is used to calculate Vr. The equation is Vr=(Vsg-vinit)/Vex per the context help for Convert Strain Gauge Reading VI.  For the strain measurement Vr is used to calculate strain using the equation in the attached picture.  As you can see, if Vsg is positive then Vr will be positive and strain will be negative for compression.  Similarly strain will be positive for tension. I would ensure that the specimen is not under some type of bending that would cause a tension on the strain gauge.

 

Best,

 

Justin P  

Justin
National Instruments
Product Support Engineer - Conditioned Measurements
0 Kudos
Message 4 of 6
(20,286 Views)

The confusion on my part is when I connect the output of the strain gauge directly to a waveform (output is in millivolts) I see negative results and believe this as accurate because it is a compression force. But when I connect the output of the strain to the convert to strain guage vi the output is positive. So I  would expect that both would be negative or both would  be positive regardless of whatever units it is displayed in. I have attached my vi to see if there is something that I am missing.

HS

0 Kudos
Message 5 of 6
(20,277 Views)

Hello, 

 

The formula used for calculating strain for the Convert Strain Gauge Reading VI is in the attached picture and has a negative sign multiplied by Vr.  So if Vr is negative then the strain will be in fact positive, this is consistant with output of the wheatstone bridge.  If the surface is under tension then the strain gauge will increase in resistance which will cause a negative voltage output from the wheatstone bridge this value has to then be converted to a positive when representing units of strain so that you get a positive strain for a strain gauge in tension.

 

Regards,

 

Justin P  

 

Justin
National Instruments
Product Support Engineer - Conditioned Measurements
0 Kudos
Message 6 of 6
(20,261 Views)