0
\$\begingroup\$

I have an experiment composed of 16 strain gages, mounted on a steal beam. The steal beam is horizontally placed under a compression cylinder for a stress test. the test took around two hours to complete and composed of different stages and levels, the maximum load achieved on the was of 160KN. The stress machine is a YLE compression machine with National instruments PXI modules for data acquisition

After retrieving the measurement results, most of the gages have stopped changing in value and become somehow saturated as shown in figure 1 Figure1 while the remaining strain gages did not preserve a linear variation, the installation used a quarter bridge and the results were calibrated at the start.

The strain gages used in this experiment is the Vishay strain gages general purpose linear module 125BT

http://www.vishaypg.com/doc?11196

So the questions are: - How did the strain gages get saturated? and how to choose the best strain gage? -what are the causing errors/noise which can cause such behavior? Does the amplifier gain limits affects the resulting values? if so why didn't it affect all of them? -what are the most common errors and noise generated when installing and using strain gages (the strain gages where soldered to a 10m shielded cable and connected to the machines inputs)

Thank you

\$\endgroup\$
  • \$\begingroup\$ Why don't you measure their ohmic value to determine if they are damaged? What is the X axis of the graph? What errors? What noise? Who knows why only some were affected - analyse the positions on the beam and draw some conclusions. \$\endgroup\$ – Andy aka Aug 28 '16 at 16:45
  • \$\begingroup\$ the X axis of the graph is the Force applied in KN, I couldn't not measure the resistance as they were directly soldered to the connection cables and it was connected to the machine, the beam was left to rest for a day to analyze it's behavior \$\endgroup\$ – chaosmind Aug 28 '16 at 16:53
1
\$\begingroup\$

enter image description here

Figure 1. Strain gauge differences.

It seems to me that there are a few problems.

  • Gauge 1 seems to be bearing the brunt of the load with 2 and 3 following. All three seem to have reached some limit and deformed permanently.
  • The traces around 4 seem to be behaving normally with one (green) exception.
  • The traces at 5 are going the opposite direction which suggests that they are in tension rather than compression.

The first and last points make it appear that the setup is mechanically faulty, that the load is not balanced between the cells and some leverage is occurring causing some gauges to go into tension. This will put further stress on the sensors in the pivot area - probably 1, 2 and 3.

\$\endgroup\$
  • \$\begingroup\$ Hello, some strain gages were placed in a way to provide both tension and compression therefore the readings have both. as the beam bent I guess the load of 157kn was directly applied to the gages 1 2 3 and 4 while others where acting in tension due to bending and some were distributed along the sides that varied a small ammount \$\endgroup\$ – chaosmind Aug 29 '16 at 15:24
  • \$\begingroup\$ Thanks for accepting my answer. I got the impression from your comment that I was on the wrong track. What did you discover? \$\endgroup\$ – Transistor Sep 2 '16 at 17:37
  • \$\begingroup\$ The middle strain gages were bearing all the load and got into an extreme tension state where they could no longer hold plus the placement was somehow wrong and the load was not centered to start with and i thought the side ones which remained constant were the damaged in addition Mark Bursic had a straight to the point as well and Scott suggestion is right as well I would choose all answers as correct! \$\endgroup\$ – chaosmind Sep 2 '16 at 17:49
  • \$\begingroup\$ Yes, multiple answers can give different viewpoints which give a better overall understanding. I often advise noobs to unaccept a quick "accept" to encourage other responses. Anyway, thanks for the update. \$\endgroup\$ – Transistor Sep 2 '16 at 18:55
0
\$\begingroup\$

Which gage series, exactly, are you using? Most of the series listed in the doc you point to have a max strain of 1.5 or 2%, but the EP series goes to +/-20%. You might try using that one, if the saturation is mechanical. If you've over-strained the devices, I don't know that they're still reliable.

A quarter bridge, I assume, is just a constant current running through your gage. If the saturation is because of the current supply reaching its voltage limit, you could try turning down the current through the strain gage.

We have no way of knowing the answers to your more detailed questions unless you provide more detailed info.

\$\endgroup\$
  • \$\begingroup\$ The used strain gage is the EA-125BT it has a 120ohm value with a +-3% I downloaded the full catalogue with the parameters "strain range" "strain level" "number of cycles" but I am looking up on what they resembles as I do not have that much knowledge \$\endgroup\$ – chaosmind Aug 29 '16 at 15:27
  • \$\begingroup\$ Looks like you have MUCH MORE than +/-3% from your figure. \$\endgroup\$ – Scott Seidman Aug 29 '16 at 16:40
  • \$\begingroup\$ Then I guess the strain handling the load directly are experiencing cracks and fatigues and are no longer useful ? \$\endgroup\$ – chaosmind Aug 29 '16 at 17:00
0
\$\begingroup\$

The strain gauges have not saturated, actualy the metal body on which the gauges were applicated has bent. According to Hook's law you entered in a region of no return, causing a plastic deformation of the body.

\$\endgroup\$
  • \$\begingroup\$ When a plastic deformation is achieved and it was achieved the load across the beam will not be distributed equally, the beam already bent at 157KN of load and i mapped the strain gages that spiked to the location they were directly under the stress cylinder the other where on the sides \$\endgroup\$ – chaosmind Aug 29 '16 at 15:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.