Trying to reduce strain gage drift

[JoeVaccar]

Hello, my name is Joe Vaccarella. I am a research assistant at the University of Rochester.

I’m currently conducting a project where I have 4 strain gages, mounted on a polycarbonate tube. (one gage on each inner face) On top of this tube is a 6x6” Polycarb plate. My current experimentation involves placing various masses on top of this plate, and then measuring the change in strain for each gage.

Hardware:

• (4) Micro-Measurement Pre-Wired Strain Gages (ID: C4A-06-235SL-120/39P)
• NI 9235 C Series Bridge Input Module
• NI CompactDAQ-9174 Chassis

I am currently able to read data from the gages, and they appear to be acting linearly for the most part. The issue I’m currently running into is that the gage measurements like to drift, seemingly in arbitrary directions.

I’ve been searching the forums for some time now, trying to find a way to reduce the drift that we’ve been encountering.

• Initially we had an idea to lower the excitation voltage, in an attempt to reduce the noise present in the system, and then possibly
  • But searching the forum, I found out that it seems that our current input module (the 9235) only supplies an excitation voltage of 2 volts, and it seems like we can’t modify it.
    • I may be wrong though and I would appreciate any feedback that can be given.

I’ve also heard that because we mounted to plastic instead of metal, we should expect increased drift, and poorer zero return.

For context, the gages will drift a few 10’s of microstrain at random points in time. This ‘shifting’ is gradual most of the time, occurring over the period of 10 secs or so. This occurs when the system is both loaded and unloaded.

Based on initial testing, we expect to encounter readings less than a 100 microstrain when we go to conduct the actual experiment, which we believe is well within the optimal working range of our current strain gages.

I would appreciate any feedback on how to reduce this drift, either through a change in something I’m doing in LabView, or a physical change to the system.

I will attach a few images for reference, along with the vi file I’ve been using.

• For context, I’ve only been using the calibration wizard as of now and doing all my testing from that window, so our current VI isn’t super relevant right now I don’t think.

Thanks for reading this!

[santo_13]

Appreciate your detailed explanation of the setup and your issue - it is very rare to see such detailed posts in NI forum.

Unfortunately, I am using LV 2019 and could not open your VI, to make it easy for the larger audience, please save the VI in LV2016 format. You can do this by opening the VI, File -> Save for Previous version... and select LabVIEW version as 16.0

To validate your theory of plastic deformation vs metals, if you got a spare strain gage, attach it to a sturdy piece of metal and observe the measurement over time, it will indicate if it is an inherent behavior of PolyCarb sheet or some electrical issue.

[JoeVaccar]

To start, thank you for the kind words Santhosh, I really appreciate it.

Now onto responding to what you said:
If I can, I'll update my original post with a correctly formatted file. I'll also attach the 2016 formatted file to this reply as well.

With regards to your suggestion to test on metal, it's very convenient as we actually already tried this. For reference, we saw reduced drift on the metal, but it was still significant. So, maybe we have an electrical issue or maybe the gages didn't bond entirely properly, we're not exactly sure.

Thanks for the help thus far, it is greatly appreciated.

Best Regards,

Joe Vaccarella

[Kevin_Price]

I know only enough about strain gage work to realize I need to rely on expert help.  In my limited experience, I've learned that for many strain gages (especially the more sensitive ones it seems), it isn't trivial to get them set up to be physically robust and give stable, reliable results.

Hopefully someone with deeper expertise can weigh in with answers, about all I can do is list some of the possible areas that need scrutiny and require more learning.  What I know (or at least *think* I know): there can be issues related to mounting / bonding, geometric configuration of multiple gages, alignment to desired axes, difficulties with the associated fine wire gages used for connections, sensitivity to temperature changes, Poisson effects from off-axis strain, and so on.  There's a *lot* of details that matter.

Inside your DAQ Assistant are a number of config choices that may need a second look.  Have you chosen the correct "Strain Configuration" corresponding to your physical setup?  Have you read up on pros and cons of different geometric configurations of strain gages?  Are you sure the gage resistance is a nominal 120 ohm and that Initial Voltage=0 is appropriate?

Sorry, I don't have the expertise to give specific advice, I only barely know some of the questions to be concerned with.

-Kevin P

P.S.  With relatively little to offer, I probably wouldn't have replied except that your name and location rang a bell.  I once worked with a Joe Vaccarella in Rochester at a since-sold-off division of Kodak, probably about 25 years ago.  Maybe your father or an uncle?   Anyway, wanted to try to help a little, meager though it may be.

[JoeVaccar]

Thank you for the tips Kevin!
I definitely agree that mounting has been a non-trivial matter. All those factors you're describing as well make a lot of sense.

With regards to your questions:
We're currently using Quarter-Bridge 1 setup, based on the capabilities of our current Input Bridge Module.
I have read up on different geometric configurations, and our group has considered using these different setups, but we have no active plans at the moment.
With regards to nominal resistance, our gages are rated for 120 ohms, and our system shoudl be running at 120, but we have not experimentally verified it.

With regards to your P.S.:
That man is definitely my grandfather, and that is such a crazy coincidence!
He's retired now, but he definitely enjoyed his Kodak days. I'll definitely tell him of this.

Thank you for all the help and your time. Have a nice day. 🙂